file.c revision 74784 
1/*
2 * file - find type of a file or files - main program.
3 *
4 * Copyright (c) Ian F. Darwin, 1987.
5 * Written by Ian F. Darwin.
6 *
7 * This software is not subject to any license of the American Telephone
8 * and Telegraph Company or of the Regents of the University of California.
9 *
10 * Permission is granted to anyone to use this software for any purpose on
11 * any computer system, and to alter it and redistribute it freely, subject
12 * to the following restrictions:
13 *
14 * 1. The author is not responsible for the consequences of use of this
15 *    software, no matter how awful, even if they arise from flaws in it.
16 *
17 * 2. The origin of this software must not be misrepresented, either by
18 *    explicit claim or by omission.  Since few users ever read sources,
19 *    credits must appear in the documentation.
20 *
21 * 3. Altered versions must be plainly marked as such, and must not be
22 *    misrepresented as being the original software.  Since few users
23 *    ever read sources, credits must appear in the documentation.
24 *
25 * 4. This notice may not be removed or altered.
26 */
27#include <stdio.h>
28#include <stdlib.h>
29#include <unistd.h>
30#include <string.h>
31#include <sys/types.h>
32#include <sys/param.h>	/* for MAXPATHLEN */
33#include <sys/stat.h>
34#include <fcntl.h>	/* for open() */
35#ifdef RESTORE_TIME
36# if (__COHERENT__ >= 0x420)
37#  include <sys/utime.h>
38# else
39#  ifdef USE_UTIMES
40#   include <sys/time.h>
41#  else
42#   include <utime.h>
43#  endif
44# endif
45#endif
46#ifdef HAVE_UNISTD_H
47#include <unistd.h>	/* for read() */
48#endif
49#ifdef HAVE_LOCALE_H
50#include <locale.h>
51#endif
52
53#include <netinet/in.h>		/* for byte swapping */
54
55#include "file.h"
56#include "patchlevel.h"
57
58#ifndef	lint
59FILE_RCSID("@(#)$Id: file.c,v 1.56 2001/03/11 20:29:16 christos Exp $")
60#endif	/* lint */
61
62
63#ifdef S_IFLNK
64# define USAGE  "Usage: %s [-bciknvzL] [-f namefile] [-m magicfiles] file...\n"
65#else
66# define USAGE  "Usage: %s [-bciknvz] [-f namefile] [-m magicfiles] file...\n"
67#endif
68
69#ifndef MAGIC
70# define MAGIC "/etc/magic"
71#endif
72
73#ifndef MAXPATHLEN
74#define	MAXPATHLEN	512
75#endif
76
77int 			/* Global command-line options 		*/
78	debug = 0, 	/* debugging 				*/
79	lflag = 0,	/* follow Symlinks (BSD only) 		*/
80	bflag = 0,	/* brief output format	 		*/
81	zflag = 0,	/* follow (uncompress) compressed files */
82	sflag = 0,	/* read block special files		*/
83	iflag = 0,
84	nobuffer = 0,   /* Do not buffer stdout */
85	kflag = 0;	/* Keep going after the first match	*/
86
87int			/* Misc globals				*/
88	nmagic = 0;	/* number of valid magic[]s 		*/
89
90struct  magic *magic;	/* array of magic entries		*/
91
92const char *magicfile;	/* where magic be found 		*/
93const char *default_magicfile = MAGIC;
94
95char *progname;		/* used throughout 			*/
96int lineno;		/* line number in the magic file	*/
97
98
99static void	unwrap		__P((char *fn));
100static void	usage		__P((void));
101#if 0
102static int	byteconv4	__P((int, int, int));
103static short	byteconv2	__P((int, int, int));
104#endif
105
106int main __P((int, char *[]));
107
108/*
109 * main - parse arguments and handle options
110 */
111int
112main(argc, argv)
113	int argc;
114	char *argv[];
115{
116	int c;
117	int action = 0, didsomefiles = 0, errflg = 0, ret = 0, app = 0;
118	char *mime;
119
120#ifdef LC_CTYPE
121	setlocale(LC_CTYPE, ""); /* makes islower etc work for other langs */
122#endif
123
124	if ((progname = strrchr(argv[0], '/')) != NULL)
125		progname++;
126	else
127		progname = argv[0];
128
129	if (!(magicfile = getenv("MAGIC")))
130		magicfile = default_magicfile;
131
132	while ((c = getopt(argc, argv, "bcdf:ikm:nsvzCL")) != EOF)
133		switch (c) {
134		case 'b':
135			++bflag;
136			break;
137		case 'c':
138			action = CHECK;
139			break;
140		case 'C':
141			action = COMPILE;
142			break;
143		case 'd':
144			++debug;
145			break;
146		case 'f':
147			if (!app) {
148				ret = apprentice(magicfile, action);
149				if (action)
150					exit(ret);
151				app = 1;
152			}
153			unwrap(optarg);
154			++didsomefiles;
155			break;
156		case 'i':
157			iflag++;
158			if ((mime = malloc(strlen(magicfile) + 5)) != NULL) {
159				(void)strcpy(mime, magicfile);
160				(void)strcat(mime, ".mime");
161				magicfile = mime;
162			}
163			break;
164		case 'k':
165			kflag = 1;
166			break;
167		case 'm':
168			magicfile = optarg;
169			break;
170		case 'n':
171			++nobuffer;
172			break;
173		case 's':
174			sflag++;
175			break;
176		case 'v':
177			(void) fprintf(stdout, "%s-%d.%d\n", progname,
178				       FILE_VERSION_MAJOR, patchlevel);
179			(void) fprintf(stdout, "magic file from %s\n",
180				       magicfile);
181			return 1;
182		case 'z':
183			zflag++;
184			break;
185#ifdef S_IFLNK
186		case 'L':
187			++lflag;
188			break;
189#endif
190		case '?':
191		default:
192			errflg++;
193			break;
194		}
195
196	if (errflg) {
197		usage();
198	}
199
200	if (!app) {
201		ret = apprentice(magicfile, action);
202		if (action)
203			exit(ret);
204		app = 1;
205	}
206
207	if (optind == argc) {
208		if (!didsomefiles) {
209			usage();
210		}
211	}
212	else {
213		int i, wid, nw;
214		for (wid = 0, i = optind; i < argc; i++) {
215			nw = strlen(argv[i]);
216			if (nw > wid)
217				wid = nw;
218		}
219		for (; optind < argc; optind++)
220			process(argv[optind], wid);
221	}
222
223	return 0;
224}
225
226
227/*
228 * unwrap -- read a file of filenames, do each one.
229 */
230static void
231unwrap(fn)
232	char *fn;
233{
234	char buf[MAXPATHLEN];
235	FILE *f;
236	int wid = 0, cwid;
237
238	if (strcmp("-", fn) == 0) {
239		f = stdin;
240		wid = 1;
241	} else {
242		if ((f = fopen(fn, "r")) == NULL) {
243			error("Cannot open `%s' (%s).\n", fn, strerror(errno));
244			/*NOTREACHED*/
245		}
246
247		while (fgets(buf, MAXPATHLEN, f) != NULL) {
248			cwid = strlen(buf) - 1;
249			if (cwid > wid)
250				wid = cwid;
251		}
252
253		rewind(f);
254	}
255
256	while (fgets(buf, MAXPATHLEN, f) != NULL) {
257		buf[strlen(buf)-1] = '\0';
258		process(buf, wid);
259		if(nobuffer)
260			(void) fflush(stdout);
261	}
262
263	(void) fclose(f);
264}
265
266
267#if 0
268/*
269 * byteconv4
270 * Input:
271 *	from		4 byte quantity to convert
272 *	same		whether to perform byte swapping
273 *	big_endian	whether we are a big endian host
274 */
275static int
276byteconv4(from, same, big_endian)
277	int from;
278	int same;
279	int big_endian;
280{
281	if (same)
282		return from;
283	else if (big_endian) {		/* lsb -> msb conversion on msb */
284		union {
285			int i;
286			char c[4];
287		} retval, tmpval;
288
289		tmpval.i = from;
290		retval.c[0] = tmpval.c[3];
291		retval.c[1] = tmpval.c[2];
292		retval.c[2] = tmpval.c[1];
293		retval.c[3] = tmpval.c[0];
294
295		return retval.i;
296	}
297	else
298		return ntohl(from);	/* msb -> lsb conversion on lsb */
299}
300
301/*
302 * byteconv2
303 * Same as byteconv4, but for shorts
304 */
305static short
306byteconv2(from, same, big_endian)
307	int from;
308	int same;
309	int big_endian;
310{
311	if (same)
312		return from;
313	else if (big_endian) {		/* lsb -> msb conversion on msb */
314		union {
315			short s;
316			char c[2];
317		} retval, tmpval;
318
319		tmpval.s = (short) from;
320		retval.c[0] = tmpval.c[1];
321		retval.c[1] = tmpval.c[0];
322
323		return retval.s;
324	}
325	else
326		return ntohs(from);	/* msb -> lsb conversion on lsb */
327}
328#endif
329
330/*
331 * process - process input file
332 */
333void
334process(inname, wid)
335	const char	*inname;
336	int wid;
337{
338	int	fd = 0;
339	static  const char stdname[] = "standard input";
340	unsigned char	buf[HOWMANY+1];	/* one extra for terminating '\0' */
341	struct stat	sb;
342	int nbytes = 0;	/* number of bytes read from a datafile */
343	char match = '\0';
344
345	if (strcmp("-", inname) == 0) {
346		if (fstat(0, &sb)<0) {
347			error("cannot fstat `%s' (%s).\n", stdname,
348			      strerror(errno));
349			/*NOTREACHED*/
350		}
351		inname = stdname;
352	}
353
354	if (wid > 0 && !bflag)
355	     (void) printf("%s:%*s ", inname,
356			   (int) (wid - strlen(inname)), "");
357
358	if (inname != stdname) {
359		/*
360		 * first try judging the file based on its filesystem status
361		 */
362		if (fsmagic(inname, &sb) != 0) {
363			putchar('\n');
364			return;
365		}
366
367		if ((fd = open(inname, O_RDONLY)) < 0) {
368			/* We can't open it, but we were able to stat it. */
369			if (sb.st_mode & 0002) ckfputs("writeable, ", stdout);
370			if (sb.st_mode & 0111) ckfputs("executable, ", stdout);
371			ckfprintf(stdout, "can't read `%s' (%s).\n",
372			    inname, strerror(errno));
373			return;
374		}
375	}
376
377
378	/*
379	 * try looking at the first HOWMANY bytes
380	 */
381	if ((nbytes = read(fd, (char *)buf, HOWMANY)) == -1) {
382		error("read failed (%s).\n", strerror(errno));
383		/*NOTREACHED*/
384	}
385
386	if (nbytes == 0)
387		ckfputs(iflag ? "application/x-empty" : "empty", stdout);
388	else {
389		buf[nbytes++] = '\0';	/* null-terminate it */
390		match = tryit(buf, nbytes, zflag);
391	}
392
393#ifdef BUILTIN_ELF
394	if (match == 's' && nbytes > 5) {
395		/*
396		 * We matched something in the file, so this *might*
397		 * be an ELF file, and the file is at least 5 bytes long,
398		 * so if it's an ELF file it has at least one byte
399		 * past the ELF magic number - try extracting information
400		 * from the ELF headers that can't easily be extracted
401		 * with rules in the magic file.
402		 */
403		tryelf(fd, buf, nbytes);
404	}
405#endif
406
407	if (inname != stdname) {
408#ifdef RESTORE_TIME
409		/*
410		 * Try to restore access, modification times if read it.
411		 * This is really *bad* because it will modify the status
412		 * time of the file... And of course this will affect
413		 * backup programs
414		 */
415# ifdef USE_UTIMES
416		struct timeval  utsbuf[2];
417		utsbuf[0].tv_sec = sb.st_atime;
418		utsbuf[1].tv_sec = sb.st_mtime;
419
420		(void) utimes(inname, utsbuf); /* don't care if loses */
421# else
422		struct utimbuf  utbuf;
423
424		utbuf.actime = sb.st_atime;
425		utbuf.modtime = sb.st_mtime;
426		(void) utime(inname, &utbuf); /* don't care if loses */
427# endif
428#endif
429		(void) close(fd);
430	}
431	(void) putchar('\n');
432}
433
434
435int
436tryit(buf, nb, zflag)
437	unsigned char *buf;
438	int nb, zflag;
439{
440	/* try compression stuff */
441	if (zflag && zmagic(buf, nb))
442		return 'z';
443
444	/* try tests in /etc/magic (or surrogate magic file) */
445	if (softmagic(buf, nb))
446		return 's';
447
448	/* try known keywords, check whether it is ASCII */
449	if (ascmagic(buf, nb))
450		return 'a';
451
452	/* abandon hope, all ye who remain here */
453	ckfputs("data", stdout);
454		return '\0';
455}
456
457static void
458usage()
459{
460	(void)fprintf(stderr, USAGE, progname);
461#ifdef QUICK
462	(void)fprintf(stderr, "Usage: %s -C [-m magic]\n", progname);
463#endif
464	exit(1);
465}
466