file.c revision 68349 
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 <string.h>
30#include <sys/types.h>
31#include <sys/param.h>	/* for MAXPATHLEN */
32#include <sys/stat.h>
33#include <fcntl.h>	/* for open() */
34#ifdef RESTORE_TIME
35# if (__COHERENT__ >= 0x420)
36#  include <sys/utime.h>
37# else
38#  ifdef USE_UTIMES
39#   include <sys/time.h>
40#  else
41#   include <utime.h>
42#  endif
43# endif
44#endif
45#ifdef HAVE_UNISTD_H
46#include <unistd.h>	/* for read() */
47#endif
48#ifdef HAVE_LOCALE_H
49#include <locale.h>
50#endif
51
52#include <netinet/in.h>		/* for byte swapping */
53
54#include "file.h"
55#include "patchlevel.h"
56
57#ifndef	lint
58FILE_RCSID("@(#)$Id: file.c,v 1.54 2000/08/05 18:30:26 christos Exp $")
59#endif	/* lint */
60
61
62#ifdef S_IFLNK
63# define USAGE  "Usage: %s [-bciknvzL] [-f namefile] [-m magicfiles] file...\n"
64#else
65# define USAGE  "Usage: %s [-bciknvz] [-f namefile] [-m magicfiles] file...\n"
66#endif
67
68#ifndef MAGIC
69# define MAGIC "/etc/magic"
70#endif
71
72#ifndef MAXPATHLEN
73#define	MAXPATHLEN	512
74#endif
75
76int 			/* Global command-line options 		*/
77	debug = 0, 	/* debugging 				*/
78	lflag = 0,	/* follow Symlinks (BSD only) 		*/
79	bflag = 0,	/* brief output format	 		*/
80	zflag = 0,	/* follow (uncompress) compressed files */
81	sflag = 0,	/* read block special files		*/
82	iflag = 0,
83	nobuffer = 0,   /* Do not buffer stdout */
84	kflag = 0;	/* Keep going after the first match	*/
85
86int			/* Misc globals				*/
87	nmagic = 0;	/* number of valid magic[]s 		*/
88
89struct  magic *magic;	/* array of magic entries		*/
90
91const char *magicfile;	/* where magic be found 		*/
92const char *default_magicfile = MAGIC;
93
94char *progname;		/* used throughout 			*/
95int lineno;		/* line number in the magic file	*/
96
97
98static void	unwrap		__P((char *fn));
99#if 0
100static int	byteconv4	__P((int, int, int));
101static short	byteconv2	__P((int, int, int));
102#endif
103
104int main __P((int, char *[]));
105
106/*
107 * main - parse arguments and handle options
108 */
109int
110main(argc, argv)
111	int argc;
112	char *argv[];
113{
114	int c;
115	int check = 0, didsomefiles = 0, errflg = 0, ret = 0, app = 0;
116	char *mime;
117
118#ifdef LC_CTYPE
119	setlocale(LC_CTYPE, ""); /* makes islower etc work for other langs */
120#endif
121
122	if ((progname = strrchr(argv[0], '/')) != NULL)
123		progname++;
124	else
125		progname = argv[0];
126
127	if (!(magicfile = getenv("MAGIC")))
128		magicfile = default_magicfile;
129
130	while ((c = getopt(argc, argv, "bcdf:ikm:nsvzL")) != EOF)
131		switch (c) {
132		case 'b':
133			++bflag;
134			break;
135		case 'c':
136			++check;
137			break;
138		case 'd':
139			++debug;
140			break;
141		case 'f':
142			if (!app) {
143				ret = apprentice(magicfile, check);
144				if (check)
145					exit(ret);
146				app = 1;
147			}
148			unwrap(optarg);
149			++didsomefiles;
150			break;
151		case 'i':
152			iflag++;
153			if ((mime = malloc(strlen(magicfile) + 5)) != NULL) {
154				(void)strcpy(mime, magicfile);
155				(void)strcat(mime, ".mime");
156				magicfile = mime;
157			}
158			break;
159		case 'k':
160			kflag = 1;
161			break;
162		case 'm':
163			magicfile = optarg;
164			break;
165		case 'n':
166			++nobuffer;
167			break;
168		case 's':
169			sflag++;
170			break;
171		case 'v':
172			(void) fprintf(stdout, "%s-%d.%d\n", progname,
173				       FILE_VERSION_MAJOR, patchlevel);
174			(void) fprintf(stdout, "magic file from %s\n",
175				       magicfile);
176			return 1;
177		case 'z':
178			zflag++;
179			break;
180#ifdef S_IFLNK
181		case 'L':
182			++lflag;
183			break;
184#endif
185		case '?':
186		default:
187			errflg++;
188			break;
189		}
190
191	if (errflg) {
192		(void) fprintf(stderr, USAGE, progname);
193		exit(2);
194	}
195
196	if (!app) {
197		ret = apprentice(magicfile, check);
198		if (check)
199			exit(ret);
200		app = 1;
201	}
202
203	if (optind == argc) {
204		if (!didsomefiles) {
205			(void)fprintf(stderr, USAGE, progname);
206			exit(2);
207		}
208	}
209	else {
210		int i, wid, nw;
211		for (wid = 0, i = optind; i < argc; i++) {
212			nw = strlen(argv[i]);
213			if (nw > wid)
214				wid = nw;
215		}
216		for (; optind < argc; optind++)
217			process(argv[optind], wid);
218	}
219
220	return 0;
221}
222
223
224/*
225 * unwrap -- read a file of filenames, do each one.
226 */
227static void
228unwrap(fn)
229	char *fn;
230{
231	char buf[MAXPATHLEN];
232	FILE *f;
233	int wid = 0, cwid;
234
235	if (strcmp("-", fn) == 0) {
236		f = stdin;
237		wid = 1;
238	} else {
239		if ((f = fopen(fn, "r")) == NULL) {
240			error("Cannot open `%s' (%s).\n", fn, strerror(errno));
241			/*NOTREACHED*/
242		}
243
244		while (fgets(buf, MAXPATHLEN, f) != NULL) {
245			cwid = strlen(buf) - 1;
246			if (cwid > wid)
247				wid = cwid;
248		}
249
250		rewind(f);
251	}
252
253	while (fgets(buf, MAXPATHLEN, f) != NULL) {
254		buf[strlen(buf)-1] = '\0';
255		process(buf, wid);
256		if(nobuffer)
257			(void) fflush(stdout);
258	}
259
260	(void) fclose(f);
261}
262
263
264#if 0
265/*
266 * byteconv4
267 * Input:
268 *	from		4 byte quantity to convert
269 *	same		whether to perform byte swapping
270 *	big_endian	whether we are a big endian host
271 */
272static int
273byteconv4(from, same, big_endian)
274	int from;
275	int same;
276	int big_endian;
277{
278	if (same)
279		return from;
280	else if (big_endian) {		/* lsb -> msb conversion on msb */
281		union {
282			int i;
283			char c[4];
284		} retval, tmpval;
285
286		tmpval.i = from;
287		retval.c[0] = tmpval.c[3];
288		retval.c[1] = tmpval.c[2];
289		retval.c[2] = tmpval.c[1];
290		retval.c[3] = tmpval.c[0];
291
292		return retval.i;
293	}
294	else
295		return ntohl(from);	/* msb -> lsb conversion on lsb */
296}
297
298/*
299 * byteconv2
300 * Same as byteconv4, but for shorts
301 */
302static short
303byteconv2(from, same, big_endian)
304	int from;
305	int same;
306	int big_endian;
307{
308	if (same)
309		return from;
310	else if (big_endian) {		/* lsb -> msb conversion on msb */
311		union {
312			short s;
313			char c[2];
314		} retval, tmpval;
315
316		tmpval.s = (short) from;
317		retval.c[0] = tmpval.c[1];
318		retval.c[1] = tmpval.c[0];
319
320		return retval.s;
321	}
322	else
323		return ntohs(from);	/* msb -> lsb conversion on lsb */
324}
325#endif
326
327/*
328 * process - process input file
329 */
330void
331process(inname, wid)
332	const char	*inname;
333	int wid;
334{
335	int	fd = 0;
336	static  const char stdname[] = "standard input";
337	unsigned char	buf[HOWMANY+1];	/* one extra for terminating '\0' */
338	struct stat	sb;
339	int nbytes = 0;	/* number of bytes read from a datafile */
340	char match = '\0';
341
342	if (strcmp("-", inname) == 0) {
343		if (fstat(0, &sb)<0) {
344			error("cannot fstat `%s' (%s).\n", stdname,
345			      strerror(errno));
346			/*NOTREACHED*/
347		}
348		inname = stdname;
349	}
350
351	if (wid > 0 && !bflag)
352	     (void) printf("%s:%*s ", inname,
353			   (int) (wid - strlen(inname)), "");
354
355	if (inname != stdname) {
356		/*
357		 * first try judging the file based on its filesystem status
358		 */
359		if (fsmagic(inname, &sb) != 0) {
360			putchar('\n');
361			return;
362		}
363
364		if ((fd = open(inname, O_RDONLY)) < 0) {
365			/* We can't open it, but we were able to stat it. */
366			if (sb.st_mode & 0002) ckfputs("writeable, ", stdout);
367			if (sb.st_mode & 0111) ckfputs("executable, ", stdout);
368			ckfprintf(stdout, "can't read `%s' (%s).\n",
369			    inname, strerror(errno));
370			return;
371		}
372	}
373
374
375	/*
376	 * try looking at the first HOWMANY bytes
377	 */
378	if ((nbytes = read(fd, (char *)buf, HOWMANY)) == -1) {
379		error("read failed (%s).\n", strerror(errno));
380		/*NOTREACHED*/
381	}
382
383	if (nbytes == 0)
384		ckfputs(iflag ? "application/x-empty" : "empty", stdout);
385	else {
386		buf[nbytes++] = '\0';	/* null-terminate it */
387		match = tryit(buf, nbytes, zflag);
388	}
389
390#ifdef BUILTIN_ELF
391	if (match == 's' && nbytes > 5) {
392		/*
393		 * We matched something in the file, so this *might*
394		 * be an ELF file, and the file is at least 5 bytes long,
395		 * so if it's an ELF file it has at least one byte
396		 * past the ELF magic number - try extracting information
397		 * from the ELF headers that can't easily be extracted
398		 * with rules in the magic file.
399		 */
400		tryelf(fd, buf, nbytes);
401	}
402#endif
403
404	if (inname != stdname) {
405#ifdef RESTORE_TIME
406		/*
407		 * Try to restore access, modification times if read it.
408		 * This is really *bad* because it will modify the status
409		 * time of the file... And of course this will affect
410		 * backup programs
411		 */
412# ifdef USE_UTIMES
413		struct timeval  utsbuf[2];
414		utsbuf[0].tv_sec = sb.st_atime;
415		utsbuf[1].tv_sec = sb.st_mtime;
416
417		(void) utimes(inname, utsbuf); /* don't care if loses */
418# else
419		struct utimbuf  utbuf;
420
421		utbuf.actime = sb.st_atime;
422		utbuf.modtime = sb.st_mtime;
423		(void) utime(inname, &utbuf); /* don't care if loses */
424# endif
425#endif
426		(void) close(fd);
427	}
428	(void) putchar('\n');
429}
430
431
432int
433tryit(buf, nb, zflag)
434	unsigned char *buf;
435	int nb, zflag;
436{
437	/* try compression stuff */
438	if (zflag && zmagic(buf, nb))
439		return 'z';
440
441	/* try tests in /etc/magic (or surrogate magic file) */
442	if (softmagic(buf, nb))
443		return 's';
444
445	/* try known keywords, check whether it is ASCII */
446	if (ascmagic(buf, nb))
447		return 'a';
448
449	/* abandon hope, all ye who remain here */
450	ckfputs("data", stdout);
451		return '\0';
452}
453