file.c revision 276415
1/*
2 * Copyright (c) Ian F. Darwin 1986-1995.
3 * Software written by Ian F. Darwin and others;
4 * maintained 1995-present by Christos Zoulas and others.
5 *
6 * Redistribution and use in source and binary forms, with or without
7 * modification, are permitted provided that the following conditions
8 * are met:
9 * 1. Redistributions of source code must retain the above copyright
10 *    notice immediately at the beginning of the file, without modification,
11 *    this list of conditions, and the following disclaimer.
12 * 2. Redistributions in binary form must reproduce the above copyright
13 *    notice, this list of conditions and the following disclaimer in the
14 *    documentation and/or other materials provided with the distribution.
15 *
16 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
17 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
18 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
19 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE FOR
20 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
21 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
22 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
23 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
24 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
25 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
26 * SUCH DAMAGE.
27 */
28/*
29 * file - find type of a file or files - main program.
30 */
31
32#include "file.h"
33
34#ifndef	lint
35FILE_RCSID("@(#)$File: file.c,v 1.159 2014/11/28 02:46:39 christos Exp $")
36#endif	/* lint */
37
38#include "magic.h"
39
40#include <stdlib.h>
41#include <unistd.h>
42#include <string.h>
43#ifdef RESTORE_TIME
44# if (__COHERENT__ >= 0x420)
45#  include <sys/utime.h>
46# else
47#  ifdef USE_UTIMES
48#   include <sys/time.h>
49#  else
50#   include <utime.h>
51#  endif
52# endif
53#endif
54#ifdef HAVE_UNISTD_H
55#include <unistd.h>	/* for read() */
56#endif
57#ifdef HAVE_WCHAR_H
58#include <wchar.h>
59#endif
60
61#if defined(HAVE_GETOPT_H) && defined(HAVE_STRUCT_OPTION)
62#include <getopt.h>
63#ifndef HAVE_GETOPT_LONG
64int getopt_long(int argc, char * const *argv, const char *optstring, const struct option *longopts, int *longindex);
65#endif
66#else
67#include "mygetopt.h"
68#endif
69
70#ifdef S_IFLNK
71#define FILE_FLAGS "-bcEhikLlNnprsvz0"
72#else
73#define FILE_FLAGS "-bcEiklNnprsvz0"
74#endif
75
76# define USAGE  \
77    "Usage: %s [" FILE_FLAGS \
78	"] [--apple] [--mime-encoding] [--mime-type]\n" \
79    "            [-e testname] [-F separator] [-f namefile] [-m magicfiles] " \
80    "file ...\n" \
81    "       %s -C [-m magicfiles]\n" \
82    "       %s [--help]\n"
83
84private int 		/* Global command-line options 		*/
85	bflag = 0,	/* brief output format	 		*/
86	nopad = 0,	/* Don't pad output			*/
87	nobuffer = 0,   /* Do not buffer stdout 		*/
88	nulsep = 0;	/* Append '\0' to the separator		*/
89
90private const char *separator = ":";	/* Default field separator	*/
91private const struct option long_options[] = {
92#define OPT(shortname, longname, opt, doc)      \
93    {longname, opt, NULL, shortname},
94#define OPT_LONGONLY(longname, opt, doc)        \
95    {longname, opt, NULL, 0},
96#include "file_opts.h"
97#undef OPT
98#undef OPT_LONGONLY
99    {0, 0, NULL, 0}
100};
101#define OPTSTRING	"bcCde:Ef:F:hiklLm:nNpP:rsvz0"
102
103private const struct {
104	const char *name;
105	int value;
106} nv[] = {
107	{ "apptype",	MAGIC_NO_CHECK_APPTYPE },
108	{ "ascii",	MAGIC_NO_CHECK_ASCII },
109	{ "cdf",	MAGIC_NO_CHECK_CDF },
110	{ "compress",	MAGIC_NO_CHECK_COMPRESS },
111	{ "elf",	MAGIC_NO_CHECK_ELF },
112	{ "encoding",	MAGIC_NO_CHECK_ENCODING },
113	{ "soft",	MAGIC_NO_CHECK_SOFT },
114	{ "tar",	MAGIC_NO_CHECK_TAR },
115	{ "text",	MAGIC_NO_CHECK_TEXT },	/* synonym for ascii */
116	{ "tokens",	MAGIC_NO_CHECK_TOKENS }, /* OBSOLETE: ignored for backwards compatibility */
117};
118
119private struct {
120	const char *name;
121	int tag;
122	size_t value;
123} pm[] = {
124	{ "indir",	MAGIC_PARAM_INDIR_MAX, 0 },
125	{ "name",	MAGIC_PARAM_NAME_MAX, 0 },
126	{ "elf_phnum",	MAGIC_PARAM_ELF_PHNUM_MAX, 0 },
127	{ "elf_shnum",	MAGIC_PARAM_ELF_SHNUM_MAX, 0 },
128};
129
130private char *progname;		/* used throughout 		*/
131
132private void usage(void);
133private void docprint(const char *);
134private void help(void);
135
136private int unwrap(struct magic_set *, const char *);
137private int process(struct magic_set *ms, const char *, int);
138private struct magic_set *load(const char *, int);
139private void setparam(const char *);
140private void applyparam(magic_t);
141
142
143/*
144 * main - parse arguments and handle options
145 */
146int
147main(int argc, char *argv[])
148{
149	int c;
150	size_t i;
151	int action = 0, didsomefiles = 0, errflg = 0;
152	int flags = 0, e = 0;
153	struct magic_set *magic = NULL;
154	int longindex;
155	const char *magicfile = NULL;		/* where the magic is	*/
156
157	/* makes islower etc work for other langs */
158#ifdef HAVE_SETLOCALE
159	(void)setlocale(LC_CTYPE, "");
160#endif
161
162#ifdef __EMX__
163	/* sh-like wildcard expansion! Shouldn't hurt at least ... */
164	_wildcard(&argc, &argv);
165#endif
166
167	if ((progname = strrchr(argv[0], '/')) != NULL)
168		progname++;
169	else
170		progname = argv[0];
171
172#ifdef S_IFLNK
173	flags |= getenv("POSIXLY_CORRECT") ? MAGIC_SYMLINK : 0;
174#endif
175	while ((c = getopt_long(argc, argv, OPTSTRING, long_options,
176	    &longindex)) != -1)
177		switch (c) {
178		case 0 :
179			switch (longindex) {
180			case 0:
181				help();
182				break;
183			case 10:
184				flags |= MAGIC_APPLE;
185				break;
186			case 11:
187				flags |= MAGIC_MIME_TYPE;
188				break;
189			case 12:
190				flags |= MAGIC_MIME_ENCODING;
191				break;
192			}
193			break;
194		case '0':
195			nulsep = 1;
196			break;
197		case 'b':
198			bflag++;
199			break;
200		case 'c':
201			action = FILE_CHECK;
202			break;
203		case 'C':
204			action = FILE_COMPILE;
205			break;
206		case 'd':
207			flags |= MAGIC_DEBUG|MAGIC_CHECK;
208			break;
209		case 'E':
210			flags |= MAGIC_ERROR;
211			break;
212		case 'e':
213			for (i = 0; i < sizeof(nv) / sizeof(nv[0]); i++)
214				if (strcmp(nv[i].name, optarg) == 0)
215					break;
216
217			if (i == sizeof(nv) / sizeof(nv[0]))
218				errflg++;
219			else
220				flags |= nv[i].value;
221			break;
222
223		case 'f':
224			if(action)
225				usage();
226			if (magic == NULL)
227				if ((magic = load(magicfile, flags)) == NULL)
228					return 1;
229			e |= unwrap(magic, optarg);
230			++didsomefiles;
231			break;
232		case 'F':
233			separator = optarg;
234			break;
235		case 'i':
236			flags |= MAGIC_MIME;
237			break;
238		case 'k':
239			flags |= MAGIC_CONTINUE;
240			break;
241		case 'l':
242			action = FILE_LIST;
243			break;
244		case 'm':
245			magicfile = optarg;
246			break;
247		case 'n':
248			++nobuffer;
249			break;
250		case 'N':
251			++nopad;
252			break;
253#if defined(HAVE_UTIME) || defined(HAVE_UTIMES)
254		case 'p':
255			flags |= MAGIC_PRESERVE_ATIME;
256			break;
257#endif
258		case 'P':
259			setparam(optarg);
260			break;
261		case 'r':
262			flags |= MAGIC_RAW;
263			break;
264			break;
265		case 's':
266			flags |= MAGIC_DEVICES;
267			break;
268		case 'v':
269			if (magicfile == NULL)
270				magicfile = magic_getpath(magicfile, action);
271			(void)fprintf(stdout, "%s-%s\n", progname, VERSION);
272			(void)fprintf(stdout, "magic file from %s\n",
273				       magicfile);
274			return 0;
275		case 'z':
276			flags |= MAGIC_COMPRESS;
277			break;
278#ifdef S_IFLNK
279		case 'L':
280			flags |= MAGIC_SYMLINK;
281			break;
282		case 'h':
283			flags &= ~MAGIC_SYMLINK;
284			break;
285#endif
286		case '?':
287		default:
288			errflg++;
289			break;
290		}
291
292	if (errflg) {
293		usage();
294	}
295	if (e)
296		return e;
297
298	if (MAGIC_VERSION != magic_version())
299		(void)fprintf(stderr, "%s: compiled magic version [%d] "
300		    "does not match with shared library magic version [%d]\n",
301		    progname, MAGIC_VERSION, magic_version());
302
303	switch(action) {
304	case FILE_CHECK:
305	case FILE_COMPILE:
306	case FILE_LIST:
307		/*
308		 * Don't try to check/compile ~/.magic unless we explicitly
309		 * ask for it.
310		 */
311		magic = magic_open(flags|MAGIC_CHECK);
312		if (magic == NULL) {
313			(void)fprintf(stderr, "%s: %s\n", progname,
314			    strerror(errno));
315			return 1;
316		}
317
318
319		switch(action) {
320		case FILE_CHECK:
321			c = magic_check(magic, magicfile);
322			break;
323		case FILE_COMPILE:
324			c = magic_compile(magic, magicfile);
325			break;
326		case FILE_LIST:
327			c = magic_list(magic, magicfile);
328			break;
329		default:
330			abort();
331		}
332		if (c == -1) {
333			(void)fprintf(stderr, "%s: %s\n", progname,
334			    magic_error(magic));
335			return 1;
336		}
337		return 0;
338	default:
339		if (magic == NULL)
340			if ((magic = load(magicfile, flags)) == NULL)
341				return 1;
342		applyparam(magic);
343	}
344
345	if (optind == argc) {
346		if (!didsomefiles)
347			usage();
348	}
349	else {
350		size_t j, wid, nw;
351		for (wid = 0, j = (size_t)optind; j < (size_t)argc; j++) {
352			nw = file_mbswidth(argv[j]);
353			if (nw > wid)
354				wid = nw;
355		}
356		/*
357		 * If bflag is only set twice, set it depending on
358		 * number of files [this is undocumented, and subject to change]
359		 */
360		if (bflag == 2) {
361			bflag = optind >= argc - 1;
362		}
363		for (; optind < argc; optind++)
364			e |= process(magic, argv[optind], wid);
365	}
366
367	if (magic)
368		magic_close(magic);
369	return e;
370}
371
372private void
373applyparam(magic_t magic)
374{
375	size_t i;
376
377	for (i = 0; i < __arraycount(pm); i++) {
378		if (pm[i].value == 0)
379			continue;
380		if (magic_setparam(magic, pm[i].tag, &pm[i].value) == -1) {
381			(void)fprintf(stderr, "%s: Can't set %s %s\n", progname,
382				pm[i].name, strerror(errno));
383			exit(1);
384		}
385	}
386}
387
388private void
389setparam(const char *p)
390{
391	size_t i;
392	char *s;
393
394	if ((s = strchr(p, '=')) == NULL)
395		goto badparm;
396
397	for (i = 0; i < __arraycount(pm); i++) {
398		if (strncmp(p, pm[i].name, s - p) != 0)
399			continue;
400		pm[i].value = atoi(s + 1);
401		return;
402	}
403badparm:
404	(void)fprintf(stderr, "%s: Unknown param %s\n", progname, p);
405	exit(1);
406}
407
408private struct magic_set *
409/*ARGSUSED*/
410load(const char *magicfile, int flags)
411{
412	struct magic_set *magic = magic_open(flags);
413	if (magic == NULL) {
414		(void)fprintf(stderr, "%s: %s\n", progname, strerror(errno));
415		return NULL;
416	}
417	if (magic_load(magic, magicfile) == -1) {
418		(void)fprintf(stderr, "%s: %s\n",
419		    progname, magic_error(magic));
420		magic_close(magic);
421		return NULL;
422	}
423	return magic;
424}
425
426/*
427 * unwrap -- read a file of filenames, do each one.
428 */
429private int
430unwrap(struct magic_set *ms, const char *fn)
431{
432	FILE *f;
433	ssize_t len;
434	char *line = NULL;
435	size_t llen = 0;
436	int wid = 0, cwid;
437	int e = 0;
438
439	if (strcmp("-", fn) == 0) {
440		f = stdin;
441		wid = 1;
442	} else {
443		if ((f = fopen(fn, "r")) == NULL) {
444			(void)fprintf(stderr, "%s: Cannot open `%s' (%s).\n",
445			    progname, fn, strerror(errno));
446			return 1;
447		}
448
449		while ((len = getline(&line, &llen, f)) > 0) {
450			if (line[len - 1] == '\n')
451				line[len - 1] = '\0';
452			cwid = file_mbswidth(line);
453			if (cwid > wid)
454				wid = cwid;
455		}
456
457		rewind(f);
458	}
459
460	while ((len = getline(&line, &llen, f)) > 0) {
461		if (line[len - 1] == '\n')
462			line[len - 1] = '\0';
463		e |= process(ms, line, wid);
464		if(nobuffer)
465			(void)fflush(stdout);
466	}
467
468	free(line);
469	(void)fclose(f);
470	return e;
471}
472
473/*
474 * Called for each input file on the command line (or in a list of files)
475 */
476private int
477process(struct magic_set *ms, const char *inname, int wid)
478{
479	const char *type;
480	int std_in = strcmp(inname, "-") == 0;
481
482	if (wid > 0 && !bflag) {
483		(void)printf("%s", std_in ? "/dev/stdin" : inname);
484		if (nulsep)
485			(void)putc('\0', stdout);
486		(void)printf("%s", separator);
487		(void)printf("%*s ",
488		    (int) (nopad ? 0 : (wid - file_mbswidth(inname))), "");
489	}
490
491	type = magic_file(ms, std_in ? NULL : inname);
492	if (type == NULL) {
493		(void)printf("ERROR: %s\n", magic_error(ms));
494		return 1;
495	} else {
496		(void)printf("%s\n", type);
497		return 0;
498	}
499}
500
501protected size_t
502file_mbswidth(const char *s)
503{
504#if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH)
505	size_t bytesconsumed, old_n, n, width = 0;
506	mbstate_t state;
507	wchar_t nextchar;
508	(void)memset(&state, 0, sizeof(mbstate_t));
509	old_n = n = strlen(s);
510
511	while (n > 0) {
512		bytesconsumed = mbrtowc(&nextchar, s, n, &state);
513		if (bytesconsumed == (size_t)(-1) ||
514		    bytesconsumed == (size_t)(-2)) {
515			/* Something went wrong, return something reasonable */
516			return old_n;
517		}
518		if (s[0] == '\n') {
519			/*
520			 * do what strlen() would do, so that caller
521			 * is always right
522			 */
523			width++;
524		} else {
525			int w = wcwidth(nextchar);
526			if (w > 0)
527				width += w;
528		}
529
530		s += bytesconsumed, n -= bytesconsumed;
531	}
532	return width;
533#else
534	return strlen(s);
535#endif
536}
537
538private void
539usage(void)
540{
541	(void)fprintf(stderr, USAGE, progname, progname, progname);
542	exit(1);
543}
544
545private void
546docprint(const char *opts)
547{
548	size_t i;
549	int comma;
550	char *sp, *p;
551
552	p = strstr(opts, "%o");
553	if (p == NULL) {
554		fprintf(stdout, "%s", opts);
555		return;
556	}
557
558	for (sp = p - 1; sp > opts && *sp == ' '; sp--)
559		continue;
560
561	fprintf(stdout, "%.*s", (int)(p - opts), opts);
562
563	comma = 0;
564	for (i = 0; i < __arraycount(nv); i++) {
565		fprintf(stdout, "%s%s", comma++ ? ", " : "", nv[i].name);
566		if (i && i % 5 == 0) {
567			fprintf(stdout, ",\n%*s", (int)(p - sp - 1), "");
568			comma = 0;
569		}
570	}
571
572	fprintf(stdout, "%s", opts + (p - opts) + 2);
573}
574
575private void
576help(void)
577{
578	(void)fputs(
579"Usage: file [OPTION...] [FILE...]\n"
580"Determine type of FILEs.\n"
581"\n", stdout);
582#define OPT(shortname, longname, opt, doc)      \
583	fprintf(stdout, "  -%c, --" longname, shortname), \
584	docprint(doc);
585#define OPT_LONGONLY(longname, opt, doc)        \
586	fprintf(stdout, "      --" longname),	\
587	docprint(doc);
588#include "file_opts.h"
589#undef OPT
590#undef OPT_LONGONLY
591	fprintf(stdout, "\nReport bugs to http://bugs.gw.com/\n");
592	exit(0);
593}
594