file.c revision 80588 
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 1.58 2001/07/22 21:04:15 christos Exp $")
60#endif	/* lint */
61
62
63#ifdef S_IFLNK
64# define USAGE  "Usage: %s [-bciknsvzL] [-f namefile] [-m magicfiles] file...\n"
65#else
66# define USAGE  "Usage: %s [-bciknsvz] [-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 = 0;	/* where the magic is		*/
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, *home, *usermagic;
119	struct stat sb;
120
121#ifdef LC_CTYPE
122	setlocale(LC_CTYPE, ""); /* makes islower etc work for other langs */
123#endif
124
125	if ((progname = strrchr(argv[0], '/')) != NULL)
126		progname++;
127	else
128		progname = argv[0];
129
130	magicfile = default_magicfile;
131	if ((usermagic = getenv("MAGIC")) != NULL)
132		magicfile = usermagic;
133	else
134		if (home = getenv("HOME")) {
135			if ((usermagic = malloc(strlen(home) + 8)) != NULL) {
136				(void)strcpy(usermagic, home);
137				(void)strcat(usermagic, "/.magic");
138				if (stat(usermagic, &sb)<0)
139					free(usermagic);
140				else
141					magicfile = usermagic;
142			}
143		}
144
145	while ((c = getopt(argc, argv, "bcdf:ikm:nsvzCL")) != EOF)
146		switch (c) {
147		case 'b':
148			++bflag;
149			break;
150		case 'c':
151			action = CHECK;
152			break;
153		case 'C':
154			action = COMPILE;
155			break;
156		case 'd':
157			++debug;
158			break;
159		case 'f':
160			if (!app) {
161				ret = apprentice(magicfile, action);
162				if (action)
163					exit(ret);
164				app = 1;
165			}
166			unwrap(optarg);
167			++didsomefiles;
168			break;
169		case 'i':
170			iflag++;
171			if ((mime = malloc(strlen(magicfile) + 6)) != NULL) {
172				(void)strcpy(mime, magicfile);
173				(void)strcat(mime, ".mime");
174				magicfile = mime;
175			}
176			break;
177		case 'k':
178			kflag = 1;
179			break;
180		case 'm':
181			magicfile = optarg;
182			break;
183		case 'n':
184			++nobuffer;
185			break;
186		case 's':
187			sflag++;
188			break;
189		case 'v':
190			(void) fprintf(stdout, "%s-%d.%d\n", progname,
191				       FILE_VERSION_MAJOR, patchlevel);
192			(void) fprintf(stdout, "magic file from %s\n",
193				       magicfile);
194			return 1;
195		case 'z':
196			zflag++;
197			break;
198#ifdef S_IFLNK
199		case 'L':
200			++lflag;
201			break;
202#endif
203		case '?':
204		default:
205			errflg++;
206			break;
207		}
208
209	if (errflg) {
210		usage();
211	}
212
213	if (!app) {
214		ret = apprentice(magicfile, action);
215		if (action)
216			exit(ret);
217		app = 1;
218	}
219
220	if (optind == argc) {
221		if (!didsomefiles) {
222			usage();
223		}
224	}
225	else {
226		int i, wid, nw;
227		for (wid = 0, i = optind; i < argc; i++) {
228			nw = strlen(argv[i]);
229			if (nw > wid)
230				wid = nw;
231		}
232		for (; optind < argc; optind++)
233			process(argv[optind], wid);
234	}
235
236	return 0;
237}
238
239
240/*
241 * unwrap -- read a file of filenames, do each one.
242 */
243static void
244unwrap(fn)
245	char *fn;
246{
247	char buf[MAXPATHLEN];
248	FILE *f;
249	int wid = 0, cwid;
250
251	if (strcmp("-", fn) == 0) {
252		f = stdin;
253		wid = 1;
254	} else {
255		if ((f = fopen(fn, "r")) == NULL) {
256			error("Cannot open `%s' (%s).\n", fn, strerror(errno));
257			/*NOTREACHED*/
258		}
259
260		while (fgets(buf, MAXPATHLEN, f) != NULL) {
261			cwid = strlen(buf) - 1;
262			if (cwid > wid)
263				wid = cwid;
264		}
265
266		rewind(f);
267	}
268
269	while (fgets(buf, MAXPATHLEN, f) != NULL) {
270		buf[strlen(buf)-1] = '\0';
271		process(buf, wid);
272		if(nobuffer)
273			(void) fflush(stdout);
274	}
275
276	(void) fclose(f);
277}
278
279
280#if 0
281/*
282 * byteconv4
283 * Input:
284 *	from		4 byte quantity to convert
285 *	same		whether to perform byte swapping
286 *	big_endian	whether we are a big endian host
287 */
288static int
289byteconv4(from, same, big_endian)
290	int from;
291	int same;
292	int big_endian;
293{
294	if (same)
295		return from;
296	else if (big_endian) {		/* lsb -> msb conversion on msb */
297		union {
298			int i;
299			char c[4];
300		} retval, tmpval;
301
302		tmpval.i = from;
303		retval.c[0] = tmpval.c[3];
304		retval.c[1] = tmpval.c[2];
305		retval.c[2] = tmpval.c[1];
306		retval.c[3] = tmpval.c[0];
307
308		return retval.i;
309	}
310	else
311		return ntohl(from);	/* msb -> lsb conversion on lsb */
312}
313
314/*
315 * byteconv2
316 * Same as byteconv4, but for shorts
317 */
318static short
319byteconv2(from, same, big_endian)
320	int from;
321	int same;
322	int big_endian;
323{
324	if (same)
325		return from;
326	else if (big_endian) {		/* lsb -> msb conversion on msb */
327		union {
328			short s;
329			char c[2];
330		} retval, tmpval;
331
332		tmpval.s = (short) from;
333		retval.c[0] = tmpval.c[1];
334		retval.c[1] = tmpval.c[0];
335
336		return retval.s;
337	}
338	else
339		return ntohs(from);	/* msb -> lsb conversion on lsb */
340}
341#endif
342
343/*
344 * process - process input file
345 */
346void
347process(inname, wid)
348	const char	*inname;
349	int wid;
350{
351	int	fd = 0;
352	static  const char stdname[] = "standard input";
353	unsigned char	buf[HOWMANY+1];	/* one extra for terminating '\0' */
354	struct stat	sb;
355	int nbytes = 0;	/* number of bytes read from a datafile */
356	char match = '\0';
357
358	if (strcmp("-", inname) == 0) {
359		if (fstat(0, &sb)<0) {
360			error("cannot fstat `%s' (%s).\n", stdname,
361			      strerror(errno));
362			/*NOTREACHED*/
363		}
364		inname = stdname;
365	}
366
367	if (wid > 0 && !bflag)
368	     (void) printf("%s:%*s ", inname,
369			   (int) (wid - strlen(inname)), "");
370
371	if (inname != stdname) {
372		/*
373		 * first try judging the file based on its filesystem status
374		 */
375		if (fsmagic(inname, &sb) != 0) {
376			putchar('\n');
377			return;
378		}
379
380		if ((fd = open(inname, O_RDONLY)) < 0) {
381			/* We can't open it, but we were able to stat it. */
382			if (sb.st_mode & 0002) ckfputs("writeable, ", stdout);
383			if (sb.st_mode & 0111) ckfputs("executable, ", stdout);
384			ckfprintf(stdout, "can't read `%s' (%s).\n",
385			    inname, strerror(errno));
386			return;
387		}
388	}
389
390
391	/*
392	 * try looking at the first HOWMANY bytes
393	 */
394	if ((nbytes = read(fd, (char *)buf, HOWMANY)) == -1) {
395		error("read failed (%s).\n", strerror(errno));
396		/*NOTREACHED*/
397	}
398
399	if (nbytes == 0)
400		ckfputs(iflag ? "application/x-empty" : "empty", stdout);
401	else {
402		buf[nbytes++] = '\0';	/* null-terminate it */
403		match = tryit(buf, nbytes, zflag);
404	}
405
406#ifdef BUILTIN_ELF
407	if (match == 's' && nbytes > 5) {
408		/*
409		 * We matched something in the file, so this *might*
410		 * be an ELF file, and the file is at least 5 bytes long,
411		 * so if it's an ELF file it has at least one byte
412		 * past the ELF magic number - try extracting information
413		 * from the ELF headers that can't easily be extracted
414		 * with rules in the magic file.
415		 */
416		tryelf(fd, buf, nbytes);
417	}
418#endif
419
420	if (inname != stdname) {
421#ifdef RESTORE_TIME
422		/*
423		 * Try to restore access, modification times if read it.
424		 * This is really *bad* because it will modify the status
425		 * time of the file... And of course this will affect
426		 * backup programs
427		 */
428# ifdef USE_UTIMES
429		struct timeval  utsbuf[2];
430		utsbuf[0].tv_sec = sb.st_atime;
431		utsbuf[1].tv_sec = sb.st_mtime;
432
433		(void) utimes(inname, utsbuf); /* don't care if loses */
434# else
435		struct utimbuf  utbuf;
436
437		utbuf.actime = sb.st_atime;
438		utbuf.modtime = sb.st_mtime;
439		(void) utime(inname, &utbuf); /* don't care if loses */
440# endif
441#endif
442		(void) close(fd);
443	}
444	(void) putchar('\n');
445}
446
447
448int
449tryit(buf, nb, zflag)
450	unsigned char *buf;
451	int nb, zflag;
452{
453	/* try compression stuff */
454	if (zflag && zmagic(buf, nb))
455		return 'z';
456
457	/* try tests in /etc/magic (or surrogate magic file) */
458	if (softmagic(buf, nb))
459		return 's';
460
461	/* try known keywords, check whether it is ASCII */
462	if (ascmagic(buf, nb))
463		return 'a';
464
465	/* abandon hope, all ye who remain here */
466	ckfputs("data", stdout);
467		return '\0';
468}
469
470static void
471usage()
472{
473	(void)fprintf(stderr, USAGE, progname);
474	(void)fprintf(stderr, "Usage: %s -C [-m magic]\n", progname);
475	exit(1);
476}
477