file.c revision 277592
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 1.160 2014/12/16 23:18:40 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	{ "elf_notes",	MAGIC_PARAM_ELF_NOTES_MAX, 0 },
129};
130
131private char *progname;		/* used throughout 		*/
132
133private void usage(void);
134private void docprint(const char *);
135private void help(void);
136
137private int unwrap(struct magic_set *, const char *);
138private int process(struct magic_set *ms, const char *, int);
139private struct magic_set *load(const char *, int);
140private void setparam(const char *);
141private void applyparam(magic_t);
142
143
144/*
145 * main - parse arguments and handle options
146 */
147int
148main(int argc, char *argv[])
149{
150	int c;
151	size_t i;
152	int action = 0, didsomefiles = 0, errflg = 0;
153	int flags = 0, e = 0;
154	struct magic_set *magic = NULL;
155	int longindex;
156	const char *magicfile = NULL;		/* where the magic is	*/
157
158	/* makes islower etc work for other langs */
159#ifdef HAVE_SETLOCALE
160	(void)setlocale(LC_CTYPE, "");
161#endif
162
163#ifdef __EMX__
164	/* sh-like wildcard expansion! Shouldn't hurt at least ... */
165	_wildcard(&argc, &argv);
166#endif
167
168	if ((progname = strrchr(argv[0], '/')) != NULL)
169		progname++;
170	else
171		progname = argv[0];
172
173#ifdef S_IFLNK
174	flags |= getenv("POSIXLY_CORRECT") ? MAGIC_SYMLINK : 0;
175#endif
176	while ((c = getopt_long(argc, argv, OPTSTRING, long_options,
177	    &longindex)) != -1)
178		switch (c) {
179		case 0 :
180			switch (longindex) {
181			case 0:
182				help();
183				break;
184			case 10:
185				flags |= MAGIC_APPLE;
186				break;
187			case 11:
188				flags |= MAGIC_MIME_TYPE;
189				break;
190			case 12:
191				flags |= MAGIC_MIME_ENCODING;
192				break;
193			}
194			break;
195		case '0':
196			nulsep = 1;
197			break;
198		case 'b':
199			bflag++;
200			break;
201		case 'c':
202			action = FILE_CHECK;
203			break;
204		case 'C':
205			action = FILE_COMPILE;
206			break;
207		case 'd':
208			flags |= MAGIC_DEBUG|MAGIC_CHECK;
209			break;
210		case 'E':
211			flags |= MAGIC_ERROR;
212			break;
213		case 'e':
214			for (i = 0; i < sizeof(nv) / sizeof(nv[0]); i++)
215				if (strcmp(nv[i].name, optarg) == 0)
216					break;
217
218			if (i == sizeof(nv) / sizeof(nv[0]))
219				errflg++;
220			else
221				flags |= nv[i].value;
222			break;
223
224		case 'f':
225			if(action)
226				usage();
227			if (magic == NULL)
228				if ((magic = load(magicfile, flags)) == NULL)
229					return 1;
230			e |= unwrap(magic, optarg);
231			++didsomefiles;
232			break;
233		case 'F':
234			separator = optarg;
235			break;
236		case 'i':
237			flags |= MAGIC_MIME;
238			break;
239		case 'k':
240			flags |= MAGIC_CONTINUE;
241			break;
242		case 'l':
243			action = FILE_LIST;
244			break;
245		case 'm':
246			magicfile = optarg;
247			break;
248		case 'n':
249			++nobuffer;
250			break;
251		case 'N':
252			++nopad;
253			break;
254#if defined(HAVE_UTIME) || defined(HAVE_UTIMES)
255		case 'p':
256			flags |= MAGIC_PRESERVE_ATIME;
257			break;
258#endif
259		case 'P':
260			setparam(optarg);
261			break;
262		case 'r':
263			flags |= MAGIC_RAW;
264			break;
265			break;
266		case 's':
267			flags |= MAGIC_DEVICES;
268			break;
269		case 'v':
270			if (magicfile == NULL)
271				magicfile = magic_getpath(magicfile, action);
272			(void)fprintf(stdout, "%s-%s\n", progname, VERSION);
273			(void)fprintf(stdout, "magic file from %s\n",
274				       magicfile);
275			return 0;
276		case 'z':
277			flags |= MAGIC_COMPRESS;
278			break;
279#ifdef S_IFLNK
280		case 'L':
281			flags |= MAGIC_SYMLINK;
282			break;
283		case 'h':
284			flags &= ~MAGIC_SYMLINK;
285			break;
286#endif
287		case '?':
288		default:
289			errflg++;
290			break;
291		}
292
293	if (errflg) {
294		usage();
295	}
296	if (e)
297		return e;
298
299	if (MAGIC_VERSION != magic_version())
300		(void)fprintf(stderr, "%s: compiled magic version [%d] "
301		    "does not match with shared library magic version [%d]\n",
302		    progname, MAGIC_VERSION, magic_version());
303
304	switch(action) {
305	case FILE_CHECK:
306	case FILE_COMPILE:
307	case FILE_LIST:
308		/*
309		 * Don't try to check/compile ~/.magic unless we explicitly
310		 * ask for it.
311		 */
312		magic = magic_open(flags|MAGIC_CHECK);
313		if (magic == NULL) {
314			(void)fprintf(stderr, "%s: %s\n", progname,
315			    strerror(errno));
316			return 1;
317		}
318
319
320		switch(action) {
321		case FILE_CHECK:
322			c = magic_check(magic, magicfile);
323			break;
324		case FILE_COMPILE:
325			c = magic_compile(magic, magicfile);
326			break;
327		case FILE_LIST:
328			c = magic_list(magic, magicfile);
329			break;
330		default:
331			abort();
332		}
333		if (c == -1) {
334			(void)fprintf(stderr, "%s: %s\n", progname,
335			    magic_error(magic));
336			return 1;
337		}
338		return 0;
339	default:
340		if (magic == NULL)
341			if ((magic = load(magicfile, flags)) == NULL)
342				return 1;
343		applyparam(magic);
344	}
345
346	if (optind == argc) {
347		if (!didsomefiles)
348			usage();
349	}
350	else {
351		size_t j, wid, nw;
352		for (wid = 0, j = (size_t)optind; j < (size_t)argc; j++) {
353			nw = file_mbswidth(argv[j]);
354			if (nw > wid)
355				wid = nw;
356		}
357		/*
358		 * If bflag is only set twice, set it depending on
359		 * number of files [this is undocumented, and subject to change]
360		 */
361		if (bflag == 2) {
362			bflag = optind >= argc - 1;
363		}
364		for (; optind < argc; optind++)
365			e |= process(magic, argv[optind], wid);
366	}
367
368	if (magic)
369		magic_close(magic);
370	return e;
371}
372
373private void
374applyparam(magic_t magic)
375{
376	size_t i;
377
378	for (i = 0; i < __arraycount(pm); i++) {
379		if (pm[i].value == 0)
380			continue;
381		if (magic_setparam(magic, pm[i].tag, &pm[i].value) == -1) {
382			(void)fprintf(stderr, "%s: Can't set %s %s\n", progname,
383				pm[i].name, strerror(errno));
384			exit(1);
385		}
386	}
387}
388
389private void
390setparam(const char *p)
391{
392	size_t i;
393	char *s;
394
395	if ((s = strchr(p, '=')) == NULL)
396		goto badparm;
397
398	for (i = 0; i < __arraycount(pm); i++) {
399		if (strncmp(p, pm[i].name, s - p) != 0)
400			continue;
401		pm[i].value = atoi(s + 1);
402		return;
403	}
404badparm:
405	(void)fprintf(stderr, "%s: Unknown param %s\n", progname, p);
406	exit(1);
407}
408
409private struct magic_set *
410/*ARGSUSED*/
411load(const char *magicfile, int flags)
412{
413	struct magic_set *magic = magic_open(flags);
414	if (magic == NULL) {
415		(void)fprintf(stderr, "%s: %s\n", progname, strerror(errno));
416		return NULL;
417	}
418	if (magic_load(magic, magicfile) == -1) {
419		(void)fprintf(stderr, "%s: %s\n",
420		    progname, magic_error(magic));
421		magic_close(magic);
422		return NULL;
423	}
424	return magic;
425}
426
427/*
428 * unwrap -- read a file of filenames, do each one.
429 */
430private int
431unwrap(struct magic_set *ms, const char *fn)
432{
433	FILE *f;
434	ssize_t len;
435	char *line = NULL;
436	size_t llen = 0;
437	int wid = 0, cwid;
438	int e = 0;
439
440	if (strcmp("-", fn) == 0) {
441		f = stdin;
442		wid = 1;
443	} else {
444		if ((f = fopen(fn, "r")) == NULL) {
445			(void)fprintf(stderr, "%s: Cannot open `%s' (%s).\n",
446			    progname, fn, strerror(errno));
447			return 1;
448		}
449
450		while ((len = getline(&line, &llen, f)) > 0) {
451			if (line[len - 1] == '\n')
452				line[len - 1] = '\0';
453			cwid = file_mbswidth(line);
454			if (cwid > wid)
455				wid = cwid;
456		}
457
458		rewind(f);
459	}
460
461	while ((len = getline(&line, &llen, f)) > 0) {
462		if (line[len - 1] == '\n')
463			line[len - 1] = '\0';
464		e |= process(ms, line, wid);
465		if(nobuffer)
466			(void)fflush(stdout);
467	}
468
469	free(line);
470	(void)fclose(f);
471	return e;
472}
473
474/*
475 * Called for each input file on the command line (or in a list of files)
476 */
477private int
478process(struct magic_set *ms, const char *inname, int wid)
479{
480	const char *type;
481	int std_in = strcmp(inname, "-") == 0;
482
483	if (wid > 0 && !bflag) {
484		(void)printf("%s", std_in ? "/dev/stdin" : inname);
485		if (nulsep)
486			(void)putc('\0', stdout);
487		(void)printf("%s", separator);
488		(void)printf("%*s ",
489		    (int) (nopad ? 0 : (wid - file_mbswidth(inname))), "");
490	}
491
492	type = magic_file(ms, std_in ? NULL : inname);
493	if (type == NULL) {
494		(void)printf("ERROR: %s\n", magic_error(ms));
495		return 1;
496	} else {
497		(void)printf("%s\n", type);
498		return 0;
499	}
500}
501
502protected size_t
503file_mbswidth(const char *s)
504{
505#if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH)
506	size_t bytesconsumed, old_n, n, width = 0;
507	mbstate_t state;
508	wchar_t nextchar;
509	(void)memset(&state, 0, sizeof(mbstate_t));
510	old_n = n = strlen(s);
511
512	while (n > 0) {
513		bytesconsumed = mbrtowc(&nextchar, s, n, &state);
514		if (bytesconsumed == (size_t)(-1) ||
515		    bytesconsumed == (size_t)(-2)) {
516			/* Something went wrong, return something reasonable */
517			return old_n;
518		}
519		if (s[0] == '\n') {
520			/*
521			 * do what strlen() would do, so that caller
522			 * is always right
523			 */
524			width++;
525		} else {
526			int w = wcwidth(nextchar);
527			if (w > 0)
528				width += w;
529		}
530
531		s += bytesconsumed, n -= bytesconsumed;
532	}
533	return width;
534#else
535	return strlen(s);
536#endif
537}
538
539private void
540usage(void)
541{
542	(void)fprintf(stderr, USAGE, progname, progname, progname);
543	exit(1);
544}
545
546private void
547docprint(const char *opts)
548{
549	size_t i;
550	int comma;
551	char *sp, *p;
552
553	p = strstr(opts, "%o");
554	if (p == NULL) {
555		fprintf(stdout, "%s", opts);
556		return;
557	}
558
559	for (sp = p - 1; sp > opts && *sp == ' '; sp--)
560		continue;
561
562	fprintf(stdout, "%.*s", (int)(p - opts), opts);
563
564	comma = 0;
565	for (i = 0; i < __arraycount(nv); i++) {
566		fprintf(stdout, "%s%s", comma++ ? ", " : "", nv[i].name);
567		if (i && i % 5 == 0) {
568			fprintf(stdout, ",\n%*s", (int)(p - sp - 1), "");
569			comma = 0;
570		}
571	}
572
573	fprintf(stdout, "%s", opts + (p - opts) + 2);
574}
575
576private void
577help(void)
578{
579	(void)fputs(
580"Usage: file [OPTION...] [FILE...]\n"
581"Determine type of FILEs.\n"
582"\n", stdout);
583#define OPT(shortname, longname, opt, doc)      \
584	fprintf(stdout, "  -%c, --" longname, shortname), \
585	docprint(doc);
586#define OPT_LONGONLY(longname, opt, doc)        \
587	fprintf(stdout, "      --" longname),	\
588	docprint(doc);
589#include "file_opts.h"
590#undef OPT
591#undef OPT_LONGONLY
592	fprintf(stdout, "\nReport bugs to http://bugs.gw.com/\n");
593	exit(0);
594}
595