file.c revision 284237
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 1.164 2015/06/03 18:21:24 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(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:rsvzZ0"
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
133#ifdef __dead
134__dead
135#endif
136private void usage(void);
137private void docprint(const char *);
138#ifdef __dead
139__dead
140#endif
141private void help(void);
142
143private int unwrap(struct magic_set *, const char *);
144private int process(struct magic_set *ms, const char *, int);
145private struct magic_set *load(const char *, int);
146private void setparam(const char *);
147private void applyparam(magic_t);
148
149
150/*
151 * main - parse arguments and handle options
152 */
153int
154main(int argc, char *argv[])
155{
156	int c;
157	size_t i;
158	int action = 0, didsomefiles = 0, errflg = 0;
159	int flags = 0, e = 0;
160	struct magic_set *magic = NULL;
161	int longindex;
162	const char *magicfile = NULL;		/* where the magic is	*/
163
164	/* makes islower etc work for other langs */
165#ifdef HAVE_SETLOCALE
166	(void)setlocale(LC_CTYPE, "");
167#endif
168
169#ifdef __EMX__
170	/* sh-like wildcard expansion! Shouldn't hurt at least ... */
171	_wildcard(&argc, &argv);
172#endif
173
174	if ((progname = strrchr(argv[0], '/')) != NULL)
175		progname++;
176	else
177		progname = argv[0];
178
179#ifdef S_IFLNK
180	flags |= getenv("POSIXLY_CORRECT") ? MAGIC_SYMLINK : 0;
181#endif
182	while ((c = getopt_long(argc, argv, OPTSTRING, long_options,
183	    &longindex)) != -1)
184		switch (c) {
185		case 0 :
186			switch (longindex) {
187			case 0:
188				help();
189				break;
190			case 10:
191				flags |= MAGIC_APPLE;
192				break;
193			case 11:
194				flags |= MAGIC_EXTENSION;
195				break;
196			case 12:
197				flags |= MAGIC_MIME_TYPE;
198				break;
199			case 13:
200				flags |= MAGIC_MIME_ENCODING;
201				break;
202			}
203			break;
204		case '0':
205			nulsep = 1;
206			break;
207		case 'b':
208			bflag++;
209			break;
210		case 'c':
211			action = FILE_CHECK;
212			break;
213		case 'C':
214			action = FILE_COMPILE;
215			break;
216		case 'd':
217			flags |= MAGIC_DEBUG|MAGIC_CHECK;
218			break;
219		case 'E':
220			flags |= MAGIC_ERROR;
221			break;
222		case 'e':
223			for (i = 0; i < sizeof(nv) / sizeof(nv[0]); i++)
224				if (strcmp(nv[i].name, optarg) == 0)
225					break;
226
227			if (i == sizeof(nv) / sizeof(nv[0]))
228				errflg++;
229			else
230				flags |= nv[i].value;
231			break;
232
233		case 'f':
234			if(action)
235				usage();
236			if (magic == NULL)
237				if ((magic = load(magicfile, flags)) == NULL)
238					return 1;
239			e |= unwrap(magic, optarg);
240			++didsomefiles;
241			break;
242		case 'F':
243			separator = optarg;
244			break;
245		case 'i':
246			flags |= MAGIC_MIME;
247			break;
248		case 'k':
249			flags |= MAGIC_CONTINUE;
250			break;
251		case 'l':
252			action = FILE_LIST;
253			break;
254		case 'm':
255			magicfile = optarg;
256			break;
257		case 'n':
258			++nobuffer;
259			break;
260		case 'N':
261			++nopad;
262			break;
263#if defined(HAVE_UTIME) || defined(HAVE_UTIMES)
264		case 'p':
265			flags |= MAGIC_PRESERVE_ATIME;
266			break;
267#endif
268		case 'P':
269			setparam(optarg);
270			break;
271		case 'r':
272			flags |= MAGIC_RAW;
273			break;
274		case 's':
275			flags |= MAGIC_DEVICES;
276			break;
277		case 'v':
278			if (magicfile == NULL)
279				magicfile = magic_getpath(magicfile, action);
280			(void)fprintf(stdout, "%s-%s\n", progname, VERSION);
281			(void)fprintf(stdout, "magic file from %s\n",
282				       magicfile);
283			return 0;
284		case 'z':
285			flags |= MAGIC_COMPRESS;
286			break;
287
288		case 'Z':
289			flags |= MAGIC_COMPRESS|MAGIC_COMPRESS_TRANSP;
290			break;
291#ifdef S_IFLNK
292		case 'L':
293			flags |= MAGIC_SYMLINK;
294			break;
295		case 'h':
296			flags &= ~MAGIC_SYMLINK;
297			break;
298#endif
299		case '?':
300		default:
301			errflg++;
302			break;
303		}
304
305	if (errflg) {
306		usage();
307	}
308	if (e)
309		return e;
310
311	if (MAGIC_VERSION != magic_version())
312		(void)fprintf(stderr, "%s: compiled magic version [%d] "
313		    "does not match with shared library magic version [%d]\n",
314		    progname, MAGIC_VERSION, magic_version());
315
316	switch(action) {
317	case FILE_CHECK:
318	case FILE_COMPILE:
319	case FILE_LIST:
320		/*
321		 * Don't try to check/compile ~/.magic unless we explicitly
322		 * ask for it.
323		 */
324		magic = magic_open(flags|MAGIC_CHECK);
325		if (magic == NULL) {
326			(void)fprintf(stderr, "%s: %s\n", progname,
327			    strerror(errno));
328			return 1;
329		}
330
331
332		switch(action) {
333		case FILE_CHECK:
334			c = magic_check(magic, magicfile);
335			break;
336		case FILE_COMPILE:
337			c = magic_compile(magic, magicfile);
338			break;
339		case FILE_LIST:
340			c = magic_list(magic, magicfile);
341			break;
342		default:
343			abort();
344		}
345		if (c == -1) {
346			(void)fprintf(stderr, "%s: %s\n", progname,
347			    magic_error(magic));
348			return 1;
349		}
350		return 0;
351	default:
352		if (magic == NULL)
353			if ((magic = load(magicfile, flags)) == NULL)
354				return 1;
355		applyparam(magic);
356	}
357
358	if (optind == argc) {
359		if (!didsomefiles)
360			usage();
361	}
362	else {
363		size_t j, wid, nw;
364		for (wid = 0, j = (size_t)optind; j < (size_t)argc; j++) {
365			nw = file_mbswidth(argv[j]);
366			if (nw > wid)
367				wid = nw;
368		}
369		/*
370		 * If bflag is only set twice, set it depending on
371		 * number of files [this is undocumented, and subject to change]
372		 */
373		if (bflag == 2) {
374			bflag = optind >= argc - 1;
375		}
376		for (; optind < argc; optind++)
377			e |= process(magic, argv[optind], wid);
378	}
379
380	if (magic)
381		magic_close(magic);
382	return e;
383}
384
385private void
386applyparam(magic_t magic)
387{
388	size_t i;
389
390	for (i = 0; i < __arraycount(pm); i++) {
391		if (pm[i].value == 0)
392			continue;
393		if (magic_setparam(magic, pm[i].tag, &pm[i].value) == -1) {
394			(void)fprintf(stderr, "%s: Can't set %s %s\n", progname,
395				pm[i].name, strerror(errno));
396			exit(1);
397		}
398	}
399}
400
401private void
402setparam(const char *p)
403{
404	size_t i;
405	char *s;
406
407	if ((s = strchr(p, '=')) == NULL)
408		goto badparm;
409
410	for (i = 0; i < __arraycount(pm); i++) {
411		if (strncmp(p, pm[i].name, s - p) != 0)
412			continue;
413		pm[i].value = atoi(s + 1);
414		return;
415	}
416badparm:
417	(void)fprintf(stderr, "%s: Unknown param %s\n", progname, p);
418	exit(1);
419}
420
421private struct magic_set *
422/*ARGSUSED*/
423load(const char *magicfile, int flags)
424{
425	struct magic_set *magic = magic_open(flags);
426	if (magic == NULL) {
427		(void)fprintf(stderr, "%s: %s\n", progname, strerror(errno));
428		return NULL;
429	}
430	if (magic_load(magic, magicfile) == -1) {
431		(void)fprintf(stderr, "%s: %s\n",
432		    progname, magic_error(magic));
433		magic_close(magic);
434		return NULL;
435	}
436	return magic;
437}
438
439/*
440 * unwrap -- read a file of filenames, do each one.
441 */
442private int
443unwrap(struct magic_set *ms, const char *fn)
444{
445	FILE *f;
446	ssize_t len;
447	char *line = NULL;
448	size_t llen = 0;
449	int wid = 0, cwid;
450	int e = 0;
451
452	if (strcmp("-", fn) == 0) {
453		f = stdin;
454		wid = 1;
455	} else {
456		if ((f = fopen(fn, "r")) == NULL) {
457			(void)fprintf(stderr, "%s: Cannot open `%s' (%s).\n",
458			    progname, fn, strerror(errno));
459			return 1;
460		}
461
462		while ((len = getline(&line, &llen, f)) > 0) {
463			if (line[len - 1] == '\n')
464				line[len - 1] = '\0';
465			cwid = file_mbswidth(line);
466			if (cwid > wid)
467				wid = cwid;
468		}
469
470		rewind(f);
471	}
472
473	while ((len = getline(&line, &llen, f)) > 0) {
474		if (line[len - 1] == '\n')
475			line[len - 1] = '\0';
476		e |= process(ms, line, wid);
477		if(nobuffer)
478			(void)fflush(stdout);
479	}
480
481	free(line);
482	(void)fclose(f);
483	return e;
484}
485
486/*
487 * Called for each input file on the command line (or in a list of files)
488 */
489private int
490process(struct magic_set *ms, const char *inname, int wid)
491{
492	const char *type;
493	int std_in = strcmp(inname, "-") == 0;
494
495	if (wid > 0 && !bflag) {
496		(void)printf("%s", std_in ? "/dev/stdin" : inname);
497		if (nulsep)
498			(void)putc('\0', stdout);
499		(void)printf("%s", separator);
500		(void)printf("%*s ",
501		    (int) (nopad ? 0 : (wid - file_mbswidth(inname))), "");
502	}
503
504	type = magic_file(ms, std_in ? NULL : inname);
505	if (type == NULL) {
506		(void)printf("ERROR: %s\n", magic_error(ms));
507		return 1;
508	} else {
509		(void)printf("%s\n", type);
510		return 0;
511	}
512}
513
514protected size_t
515file_mbswidth(const char *s)
516{
517#if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH)
518	size_t bytesconsumed, old_n, n, width = 0;
519	mbstate_t state;
520	wchar_t nextchar;
521	(void)memset(&state, 0, sizeof(mbstate_t));
522	old_n = n = strlen(s);
523
524	while (n > 0) {
525		bytesconsumed = mbrtowc(&nextchar, s, n, &state);
526		if (bytesconsumed == (size_t)(-1) ||
527		    bytesconsumed == (size_t)(-2)) {
528			/* Something went wrong, return something reasonable */
529			return old_n;
530		}
531		if (s[0] == '\n') {
532			/*
533			 * do what strlen() would do, so that caller
534			 * is always right
535			 */
536			width++;
537		} else {
538			int w = wcwidth(nextchar);
539			if (w > 0)
540				width += w;
541		}
542
543		s += bytesconsumed, n -= bytesconsumed;
544	}
545	return width;
546#else
547	return strlen(s);
548#endif
549}
550
551private void
552usage(void)
553{
554	(void)fprintf(stderr, USAGE, progname, progname, progname);
555	exit(1);
556}
557
558private void
559docprint(const char *opts)
560{
561	size_t i;
562	int comma;
563	char *sp, *p;
564
565	p = strstr(opts, "%o");
566	if (p == NULL) {
567		fprintf(stdout, "%s", opts);
568		return;
569	}
570
571	for (sp = p - 1; sp > opts && *sp == ' '; sp--)
572		continue;
573
574	fprintf(stdout, "%.*s", (int)(p - opts), opts);
575
576	comma = 0;
577	for (i = 0; i < __arraycount(nv); i++) {
578		fprintf(stdout, "%s%s", comma++ ? ", " : "", nv[i].name);
579		if (i && i % 5 == 0) {
580			fprintf(stdout, ",\n%*s", (int)(p - sp - 1), "");
581			comma = 0;
582		}
583	}
584
585	fprintf(stdout, "%s", opts + (p - opts) + 2);
586}
587
588private void
589help(void)
590{
591	(void)fputs(
592"Usage: file [OPTION...] [FILE...]\n"
593"Determine type of FILEs.\n"
594"\n", stdout);
595#define OPT(shortname, longname, opt, doc)      \
596	fprintf(stdout, "  -%c, --" longname, shortname), \
597	docprint(doc);
598#define OPT_LONGONLY(longname, opt, doc)        \
599	fprintf(stdout, "      --" longname),	\
600	docprint(doc);
601#include "file_opts.h"
602#undef OPT
603#undef OPT_LONGONLY
604	fprintf(stdout, "\nReport bugs to http://bugs.gw.com/\n");
605	exit(0);
606}
607