file.c revision 309848
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 1.172 2016/10/24 15:21:07 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 "-bcEhikLlNnprsvzZ0"
72#else
73#define FILE_FLAGS "-bcEiklNnprsvzZ0"
74#endif
75
76# define USAGE  \
77    "Usage: %s [" FILE_FLAGS \
78	"] [--apple] [--extension] [--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_HELP		1
93#define OPT_APPLE		2
94#define OPT_EXTENSIONS		3
95#define OPT_MIME_TYPE		4
96#define OPT_MIME_ENCODING	5
97#define OPT(shortname, longname, opt, def, doc)      \
98    {longname, opt, NULL, shortname},
99#define OPT_LONGONLY(longname, opt, def, doc, id)        \
100    {longname, opt, NULL, id},
101#include "file_opts.h"
102#undef OPT
103#undef OPT_LONGONLY
104    {0, 0, NULL, 0}
105};
106#define OPTSTRING	"bcCde:Ef:F:hiklLm:nNpP:rsvzZ0"
107
108private const struct {
109	const char *name;
110	int value;
111} nv[] = {
112	{ "apptype",	MAGIC_NO_CHECK_APPTYPE },
113	{ "ascii",	MAGIC_NO_CHECK_ASCII },
114	{ "cdf",	MAGIC_NO_CHECK_CDF },
115	{ "compress",	MAGIC_NO_CHECK_COMPRESS },
116	{ "elf",	MAGIC_NO_CHECK_ELF },
117	{ "encoding",	MAGIC_NO_CHECK_ENCODING },
118	{ "soft",	MAGIC_NO_CHECK_SOFT },
119	{ "tar",	MAGIC_NO_CHECK_TAR },
120	{ "text",	MAGIC_NO_CHECK_TEXT },	/* synonym for ascii */
121	{ "tokens",	MAGIC_NO_CHECK_TOKENS }, /* OBSOLETE: ignored for backwards compatibility */
122};
123
124private struct {
125	const char *name;
126	int tag;
127	size_t value;
128} pm[] = {
129	{ "indir",	MAGIC_PARAM_INDIR_MAX, 0 },
130	{ "name",	MAGIC_PARAM_NAME_MAX, 0 },
131	{ "elf_phnum",	MAGIC_PARAM_ELF_PHNUM_MAX, 0 },
132	{ "elf_shnum",	MAGIC_PARAM_ELF_SHNUM_MAX, 0 },
133	{ "elf_notes",	MAGIC_PARAM_ELF_NOTES_MAX, 0 },
134	{ "regex",	MAGIC_PARAM_REGEX_MAX, 0 },
135	{ "bytes",	MAGIC_PARAM_BYTES_MAX, 0 },
136};
137
138private char *progname;		/* used throughout 		*/
139private int posixly;
140
141#ifdef __dead
142__dead
143#endif
144private void usage(void);
145private void docprint(const char *, int);
146#ifdef __dead
147__dead
148#endif
149private void help(void);
150
151private int unwrap(struct magic_set *, const char *);
152private int process(struct magic_set *ms, const char *, int);
153private struct magic_set *load(const char *, int);
154private void setparam(const char *);
155private void applyparam(magic_t);
156
157
158/*
159 * main - parse arguments and handle options
160 */
161int
162main(int argc, char *argv[])
163{
164	int c;
165	size_t i;
166	int action = 0, didsomefiles = 0, errflg = 0;
167	int flags = 0, e = 0;
168	struct magic_set *magic = NULL;
169	int longindex;
170	const char *magicfile = NULL;		/* where the magic is	*/
171
172	/* makes islower etc work for other langs */
173#ifdef HAVE_SETLOCALE
174	(void)setlocale(LC_CTYPE, "");
175#endif
176
177#ifdef __EMX__
178	/* sh-like wildcard expansion! Shouldn't hurt at least ... */
179	_wildcard(&argc, &argv);
180#endif
181
182	if ((progname = strrchr(argv[0], '/')) != NULL)
183		progname++;
184	else
185		progname = argv[0];
186
187#ifdef S_IFLNK
188	posixly = getenv("POSIXLY_CORRECT") != NULL;
189	flags |=  posixly ? MAGIC_SYMLINK : 0;
190#endif
191	while ((c = getopt_long(argc, argv, OPTSTRING, long_options,
192	    &longindex)) != -1)
193		switch (c) {
194		case OPT_HELP:
195			help();
196			break;
197		case OPT_APPLE:
198			flags |= MAGIC_APPLE;
199			break;
200		case OPT_EXTENSIONS:
201			flags |= MAGIC_EXTENSION;
202			break;
203		case OPT_MIME_TYPE:
204			flags |= MAGIC_MIME_TYPE;
205			break;
206		case OPT_MIME_ENCODING:
207			flags |= MAGIC_MIME_ENCODING;
208			break;
209		case '0':
210			nulsep++;
211			break;
212		case 'b':
213			bflag++;
214			break;
215		case 'c':
216			action = FILE_CHECK;
217			break;
218		case 'C':
219			action = FILE_COMPILE;
220			break;
221		case 'd':
222			flags |= MAGIC_DEBUG|MAGIC_CHECK;
223			break;
224		case 'E':
225			flags |= MAGIC_ERROR;
226			break;
227		case 'e':
228			for (i = 0; i < sizeof(nv) / sizeof(nv[0]); i++)
229				if (strcmp(nv[i].name, optarg) == 0)
230					break;
231
232			if (i == sizeof(nv) / sizeof(nv[0]))
233				errflg++;
234			else
235				flags |= nv[i].value;
236			break;
237
238		case 'f':
239			if(action)
240				usage();
241			if (magic == NULL)
242				if ((magic = load(magicfile, flags)) == NULL)
243					return 1;
244			applyparam(magic);
245			e |= unwrap(magic, optarg);
246			++didsomefiles;
247			break;
248		case 'F':
249			separator = optarg;
250			break;
251		case 'i':
252			flags |= MAGIC_MIME;
253			break;
254		case 'k':
255			flags |= MAGIC_CONTINUE;
256			break;
257		case 'l':
258			action = FILE_LIST;
259			break;
260		case 'm':
261			magicfile = optarg;
262			break;
263		case 'n':
264			++nobuffer;
265			break;
266		case 'N':
267			++nopad;
268			break;
269#if defined(HAVE_UTIME) || defined(HAVE_UTIMES)
270		case 'p':
271			flags |= MAGIC_PRESERVE_ATIME;
272			break;
273#endif
274		case 'P':
275			setparam(optarg);
276			break;
277		case 'r':
278			flags |= MAGIC_RAW;
279			break;
280		case 's':
281			flags |= MAGIC_DEVICES;
282			break;
283		case 'v':
284			if (magicfile == NULL)
285				magicfile = magic_getpath(magicfile, action);
286			(void)fprintf(stdout, "%s-%s\n", progname, VERSION);
287			(void)fprintf(stdout, "magic file from %s\n",
288				       magicfile);
289			return 0;
290		case 'z':
291			flags |= MAGIC_COMPRESS;
292			break;
293
294		case 'Z':
295			flags |= MAGIC_COMPRESS|MAGIC_COMPRESS_TRANSP;
296			break;
297#ifdef S_IFLNK
298		case 'L':
299			flags |= MAGIC_SYMLINK;
300			break;
301		case 'h':
302			flags &= ~MAGIC_SYMLINK;
303			break;
304#endif
305		case '?':
306		default:
307			errflg++;
308			break;
309		}
310
311	if (errflg) {
312		usage();
313	}
314	if (e)
315		return e;
316
317	if (MAGIC_VERSION != magic_version())
318		(void)fprintf(stderr, "%s: compiled magic version [%d] "
319		    "does not match with shared library magic version [%d]\n",
320		    progname, MAGIC_VERSION, magic_version());
321
322	switch(action) {
323	case FILE_CHECK:
324	case FILE_COMPILE:
325	case FILE_LIST:
326		/*
327		 * Don't try to check/compile ~/.magic unless we explicitly
328		 * ask for it.
329		 */
330		magic = magic_open(flags|MAGIC_CHECK);
331		if (magic == NULL) {
332			(void)fprintf(stderr, "%s: %s\n", progname,
333			    strerror(errno));
334			return 1;
335		}
336
337
338		switch(action) {
339		case FILE_CHECK:
340			c = magic_check(magic, magicfile);
341			break;
342		case FILE_COMPILE:
343			c = magic_compile(magic, magicfile);
344			break;
345		case FILE_LIST:
346			c = magic_list(magic, magicfile);
347			break;
348		default:
349			abort();
350		}
351		if (c == -1) {
352			(void)fprintf(stderr, "%s: %s\n", progname,
353			    magic_error(magic));
354			e = 1;
355			goto out;
356		}
357		goto out;
358	default:
359		if (magic == NULL)
360			if ((magic = load(magicfile, flags)) == NULL)
361				return 1;
362		applyparam(magic);
363	}
364
365	if (optind == argc) {
366		if (!didsomefiles)
367			usage();
368	}
369	else {
370		size_t j, wid, nw;
371		for (wid = 0, j = (size_t)optind; j < (size_t)argc; j++) {
372			nw = file_mbswidth(argv[j]);
373			if (nw > wid)
374				wid = nw;
375		}
376		/*
377		 * If bflag is only set twice, set it depending on
378		 * number of files [this is undocumented, and subject to change]
379		 */
380		if (bflag == 2) {
381			bflag = optind >= argc - 1;
382		}
383		for (; optind < argc; optind++)
384			e |= process(magic, argv[optind], wid);
385	}
386
387out:
388	if (magic)
389		magic_close(magic);
390	return e;
391}
392
393private void
394applyparam(magic_t magic)
395{
396	size_t i;
397
398	for (i = 0; i < __arraycount(pm); i++) {
399		if (pm[i].value == 0)
400			continue;
401		if (magic_setparam(magic, pm[i].tag, &pm[i].value) == -1) {
402			(void)fprintf(stderr, "%s: Can't set %s %s\n", progname,
403				pm[i].name, strerror(errno));
404			exit(1);
405		}
406	}
407}
408
409private void
410setparam(const char *p)
411{
412	size_t i;
413	char *s;
414
415	if ((s = strchr(p, '=')) == NULL)
416		goto badparm;
417
418	for (i = 0; i < __arraycount(pm); i++) {
419		if (strncmp(p, pm[i].name, s - p) != 0)
420			continue;
421		pm[i].value = atoi(s + 1);
422		return;
423	}
424badparm:
425	(void)fprintf(stderr, "%s: Unknown param %s\n", progname, p);
426	exit(1);
427}
428
429private struct magic_set *
430/*ARGSUSED*/
431load(const char *magicfile, int flags)
432{
433	struct magic_set *magic = magic_open(flags);
434	const char *e;
435
436	if (magic == NULL) {
437		(void)fprintf(stderr, "%s: %s\n", progname, strerror(errno));
438		return NULL;
439	}
440	if (magic_load(magic, magicfile) == -1) {
441		(void)fprintf(stderr, "%s: %s\n",
442		    progname, magic_error(magic));
443		magic_close(magic);
444		return NULL;
445	}
446	if ((e = magic_error(magic)) != NULL)
447		(void)fprintf(stderr, "%s: Warning: %s\n", progname, e);
448	return magic;
449}
450
451/*
452 * unwrap -- read a file of filenames, do each one.
453 */
454private int
455unwrap(struct magic_set *ms, const char *fn)
456{
457	FILE *f;
458	ssize_t len;
459	char *line = NULL;
460	size_t llen = 0;
461	int wid = 0, cwid;
462	int e = 0;
463
464	if (strcmp("-", fn) == 0) {
465		f = stdin;
466		wid = 1;
467	} else {
468		if ((f = fopen(fn, "r")) == NULL) {
469			(void)fprintf(stderr, "%s: Cannot open `%s' (%s).\n",
470			    progname, fn, strerror(errno));
471			return 1;
472		}
473
474		while ((len = getline(&line, &llen, f)) > 0) {
475			if (line[len - 1] == '\n')
476				line[len - 1] = '\0';
477			cwid = file_mbswidth(line);
478			if (cwid > wid)
479				wid = cwid;
480		}
481
482		rewind(f);
483	}
484
485	while ((len = getline(&line, &llen, f)) > 0) {
486		if (line[len - 1] == '\n')
487			line[len - 1] = '\0';
488		e |= process(ms, line, wid);
489		if(nobuffer)
490			(void)fflush(stdout);
491	}
492
493	free(line);
494	(void)fclose(f);
495	return e;
496}
497
498/*
499 * Called for each input file on the command line (or in a list of files)
500 */
501private int
502process(struct magic_set *ms, const char *inname, int wid)
503{
504	const char *type, c = nulsep > 1 ? '\0' : '\n';
505	int std_in = strcmp(inname, "-") == 0;
506
507	if (wid > 0 && !bflag) {
508		(void)printf("%s", std_in ? "/dev/stdin" : inname);
509		if (nulsep)
510			(void)putc('\0', stdout);
511		if (nulsep < 2) {
512			(void)printf("%s", separator);
513			(void)printf("%*s ",
514			    (int) (nopad ? 0 : (wid - file_mbswidth(inname))),
515			    "");
516		}
517	}
518
519	type = magic_file(ms, std_in ? NULL : inname);
520
521	if (type == NULL) {
522		(void)printf("ERROR: %s%c", magic_error(ms), c);
523		return 1;
524	} else {
525		(void)printf("%s%c", type, c);
526		return 0;
527	}
528}
529
530protected size_t
531file_mbswidth(const char *s)
532{
533#if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH)
534	size_t bytesconsumed, old_n, n, width = 0;
535	mbstate_t state;
536	wchar_t nextchar;
537	(void)memset(&state, 0, sizeof(mbstate_t));
538	old_n = n = strlen(s);
539
540	while (n > 0) {
541		bytesconsumed = mbrtowc(&nextchar, s, n, &state);
542		if (bytesconsumed == (size_t)(-1) ||
543		    bytesconsumed == (size_t)(-2)) {
544			/* Something went wrong, return something reasonable */
545			return old_n;
546		}
547		if (s[0] == '\n') {
548			/*
549			 * do what strlen() would do, so that caller
550			 * is always right
551			 */
552			width++;
553		} else {
554			int w = wcwidth(nextchar);
555			if (w > 0)
556				width += w;
557		}
558
559		s += bytesconsumed, n -= bytesconsumed;
560	}
561	return width;
562#else
563	return strlen(s);
564#endif
565}
566
567private void
568usage(void)
569{
570	(void)fprintf(stderr, USAGE, progname, progname, progname);
571	exit(1);
572}
573
574private void
575defprint(int def)
576{
577	if (!def)
578		return;
579	if (((def & 1) && posixly) || ((def & 2) && !posixly))
580		fprintf(stdout, " (default)");
581	fputc('\n', stdout);
582}
583
584private void
585docprint(const char *opts, int def)
586{
587	size_t i;
588	int comma;
589	char *sp, *p;
590
591	p = strstr(opts, "%o");
592	if (p == NULL) {
593		fprintf(stdout, "%s", opts);
594		defprint(def);
595		return;
596	}
597
598	for (sp = p - 1; sp > opts && *sp == ' '; sp--)
599		continue;
600
601	fprintf(stdout, "%.*s", (int)(p - opts), opts);
602
603	comma = 0;
604	for (i = 0; i < __arraycount(nv); i++) {
605		fprintf(stdout, "%s%s", comma++ ? ", " : "", nv[i].name);
606		if (i && i % 5 == 0) {
607			fprintf(stdout, ",\n%*s", (int)(p - sp - 1), "");
608			comma = 0;
609		}
610	}
611
612	fprintf(stdout, "%s", opts + (p - opts) + 2);
613}
614
615private void
616help(void)
617{
618	(void)fputs(
619"Usage: file [OPTION...] [FILE...]\n"
620"Determine type of FILEs.\n"
621"\n", stdout);
622#define OPT(shortname, longname, opt, def, doc)      \
623	fprintf(stdout, "  -%c, --" longname, shortname), \
624	docprint(doc, def);
625#define OPT_LONGONLY(longname, opt, def, doc, id)        \
626	fprintf(stdout, "      --" longname),	\
627	docprint(doc, def);
628#include "file_opts.h"
629#undef OPT
630#undef OPT_LONGONLY
631	fprintf(stdout, "\nReport bugs to http://bugs.gw.com/\n");
632	exit(0);
633}
634