file.c revision 267897
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 1.153 2014/02/11 15:41:04 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_LOCALE_H
58#include <locale.h>
59#endif
60#ifdef HAVE_WCHAR_H
61#include <wchar.h>
62#endif
63
64#if defined(HAVE_GETOPT_H) && defined(HAVE_STRUCT_OPTION)
65#include <getopt.h>
66#ifndef HAVE_GETOPT_LONG
67int getopt_long(int argc, char * const *argv, const char *optstring, const struct option *longopts, int *longindex);
68#endif
69#else
70#include "mygetopt.h"
71#endif
72
73#ifdef S_IFLNK
74#define FILE_FLAGS "-bcEhikLlNnprsvz0"
75#else
76#define FILE_FLAGS "-bcEiklNnprsvz0"
77#endif
78
79# define USAGE  \
80    "Usage: %s [" FILE_FLAGS \
81	"] [--apple] [--mime-encoding] [--mime-type]\n" \
82    "            [-e testname] [-F separator] [-f namefile] [-m magicfiles] " \
83    "file ...\n" \
84    "       %s -C [-m magicfiles]\n" \
85    "       %s [--help]\n"
86
87private int 		/* Global command-line options 		*/
88	bflag = 0,	/* brief output format	 		*/
89	nopad = 0,	/* Don't pad output			*/
90	nobuffer = 0,   /* Do not buffer stdout 		*/
91	nulsep = 0;	/* Append '\0' to the separator		*/
92
93private const char *separator = ":";	/* Default field separator	*/
94private const struct option long_options[] = {
95#define OPT(shortname, longname, opt, doc)      \
96    {longname, opt, NULL, shortname},
97#define OPT_LONGONLY(longname, opt, doc)        \
98    {longname, opt, NULL, 0},
99#include "file_opts.h"
100#undef OPT
101#undef OPT_LONGONLY
102    {0, 0, NULL, 0}
103};
104#define OPTSTRING	"bcCde:Ef:F:hiklLm:nNprsvz0"
105
106private const struct {
107	const char *name;
108	int value;
109} nv[] = {
110	{ "apptype",	MAGIC_NO_CHECK_APPTYPE },
111	{ "ascii",	MAGIC_NO_CHECK_ASCII },
112	{ "cdf",	MAGIC_NO_CHECK_CDF },
113	{ "compress",	MAGIC_NO_CHECK_COMPRESS },
114	{ "elf",	MAGIC_NO_CHECK_ELF },
115	{ "encoding",	MAGIC_NO_CHECK_ENCODING },
116	{ "soft",	MAGIC_NO_CHECK_SOFT },
117	{ "tar",	MAGIC_NO_CHECK_TAR },
118	{ "text",	MAGIC_NO_CHECK_TEXT },	/* synonym for ascii */
119	{ "tokens",	MAGIC_NO_CHECK_TOKENS }, /* OBSOLETE: ignored for backwards compatibility */
120};
121
122private char *progname;		/* used throughout 		*/
123
124private void usage(void);
125private void docprint(const char *);
126private void help(void);
127
128private int unwrap(struct magic_set *, const char *);
129private int process(struct magic_set *ms, const char *, int);
130private struct magic_set *load(const char *, int);
131
132
133/*
134 * main - parse arguments and handle options
135 */
136int
137main(int argc, char *argv[])
138{
139	int c;
140	size_t i;
141	int action = 0, didsomefiles = 0, errflg = 0;
142	int flags = 0, e = 0;
143	struct magic_set *magic = NULL;
144	int longindex;
145	const char *magicfile = NULL;		/* where the magic is	*/
146
147	/* makes islower etc work for other langs */
148	(void)setlocale(LC_CTYPE, "");
149
150#ifdef __EMX__
151	/* sh-like wildcard expansion! Shouldn't hurt at least ... */
152	_wildcard(&argc, &argv);
153#endif
154
155	if ((progname = strrchr(argv[0], '/')) != NULL)
156		progname++;
157	else
158		progname = argv[0];
159
160#ifdef S_IFLNK
161	flags |= getenv("POSIXLY_CORRECT") ? MAGIC_SYMLINK : 0;
162#endif
163	while ((c = getopt_long(argc, argv, OPTSTRING, long_options,
164	    &longindex)) != -1)
165		switch (c) {
166		case 0 :
167			switch (longindex) {
168			case 0:
169				help();
170				break;
171			case 10:
172				flags |= MAGIC_APPLE;
173				break;
174			case 11:
175				flags |= MAGIC_MIME_TYPE;
176				break;
177			case 12:
178				flags |= MAGIC_MIME_ENCODING;
179				break;
180			}
181			break;
182		case '0':
183			nulsep = 1;
184			break;
185		case 'b':
186			bflag++;
187			break;
188		case 'c':
189			action = FILE_CHECK;
190			break;
191		case 'C':
192			action = FILE_COMPILE;
193			break;
194		case 'd':
195			flags |= MAGIC_DEBUG|MAGIC_CHECK;
196			break;
197		case 'E':
198			flags |= MAGIC_ERROR;
199			break;
200		case 'e':
201			for (i = 0; i < sizeof(nv) / sizeof(nv[0]); i++)
202				if (strcmp(nv[i].name, optarg) == 0)
203					break;
204
205			if (i == sizeof(nv) / sizeof(nv[0]))
206				errflg++;
207			else
208				flags |= nv[i].value;
209			break;
210
211		case 'f':
212			if(action)
213				usage();
214			if (magic == NULL)
215				if ((magic = load(magicfile, flags)) == NULL)
216					return 1;
217			e |= unwrap(magic, optarg);
218			++didsomefiles;
219			break;
220		case 'F':
221			separator = optarg;
222			break;
223		case 'i':
224			flags |= MAGIC_MIME;
225			break;
226		case 'k':
227			flags |= MAGIC_CONTINUE;
228			break;
229		case 'l':
230			action = FILE_LIST;
231			break;
232		case 'm':
233			magicfile = optarg;
234			break;
235		case 'n':
236			++nobuffer;
237			break;
238		case 'N':
239			++nopad;
240			break;
241#if defined(HAVE_UTIME) || defined(HAVE_UTIMES)
242		case 'p':
243			flags |= MAGIC_PRESERVE_ATIME;
244			break;
245#endif
246		case 'r':
247			flags |= MAGIC_RAW;
248			break;
249		case 's':
250			flags |= MAGIC_DEVICES;
251			break;
252		case 'v':
253			if (magicfile == NULL)
254				magicfile = magic_getpath(magicfile, action);
255			(void)fprintf(stdout, "%s-%s\n", progname, VERSION);
256			(void)fprintf(stdout, "magic file from %s\n",
257				       magicfile);
258			return 0;
259		case 'z':
260			flags |= MAGIC_COMPRESS;
261			break;
262#ifdef S_IFLNK
263		case 'L':
264			flags |= MAGIC_SYMLINK;
265			break;
266		case 'h':
267			flags &= ~MAGIC_SYMLINK;
268			break;
269#endif
270		case '?':
271		default:
272			errflg++;
273			break;
274		}
275
276	if (errflg) {
277		usage();
278	}
279	if (e)
280		return e;
281
282	if (MAGIC_VERSION != magic_version())
283		(void)fprintf(stderr, "%s: compiled magic version [%d] "
284		    "does not match with shared library magic version [%d]\n",
285		    progname, MAGIC_VERSION, magic_version());
286
287	switch(action) {
288	case FILE_CHECK:
289	case FILE_COMPILE:
290	case FILE_LIST:
291		/*
292		 * Don't try to check/compile ~/.magic unless we explicitly
293		 * ask for it.
294		 */
295		magic = magic_open(flags|MAGIC_CHECK);
296		if (magic == NULL) {
297			(void)fprintf(stderr, "%s: %s\n", progname,
298			    strerror(errno));
299			return 1;
300		}
301		switch(action) {
302		case FILE_CHECK:
303			c = magic_check(magic, magicfile);
304			break;
305		case FILE_COMPILE:
306			c = magic_compile(magic, magicfile);
307			break;
308		case FILE_LIST:
309			c = magic_list(magic, magicfile);
310			break;
311		default:
312			abort();
313		}
314		if (c == -1) {
315			(void)fprintf(stderr, "%s: %s\n", progname,
316			    magic_error(magic));
317			return 1;
318		}
319		return 0;
320	default:
321		if (magic == NULL)
322			if ((magic = load(magicfile, flags)) == NULL)
323				return 1;
324		break;
325	}
326
327	if (optind == argc) {
328		if (!didsomefiles)
329			usage();
330	}
331	else {
332		size_t j, wid, nw;
333		for (wid = 0, j = (size_t)optind; j < (size_t)argc; j++) {
334			nw = file_mbswidth(argv[j]);
335			if (nw > wid)
336				wid = nw;
337		}
338		/*
339		 * If bflag is only set twice, set it depending on
340		 * number of files [this is undocumented, and subject to change]
341		 */
342		if (bflag == 2) {
343			bflag = optind >= argc - 1;
344		}
345		for (; optind < argc; optind++)
346			e |= process(magic, argv[optind], wid);
347	}
348
349	if (magic)
350		magic_close(magic);
351	return e;
352}
353
354
355private struct magic_set *
356/*ARGSUSED*/
357load(const char *magicfile, int flags)
358{
359	struct magic_set *magic = magic_open(flags);
360	if (magic == NULL) {
361		(void)fprintf(stderr, "%s: %s\n", progname, strerror(errno));
362		return NULL;
363	}
364	if (magic_load(magic, magicfile) == -1) {
365		(void)fprintf(stderr, "%s: %s\n",
366		    progname, magic_error(magic));
367		magic_close(magic);
368		return NULL;
369	}
370	return magic;
371}
372
373/*
374 * unwrap -- read a file of filenames, do each one.
375 */
376private int
377unwrap(struct magic_set *ms, const char *fn)
378{
379	FILE *f;
380	ssize_t len;
381	char *line = NULL;
382	size_t llen = 0;
383	int wid = 0, cwid;
384	int e = 0;
385
386	if (strcmp("-", fn) == 0) {
387		f = stdin;
388		wid = 1;
389	} else {
390		if ((f = fopen(fn, "r")) == NULL) {
391			(void)fprintf(stderr, "%s: Cannot open `%s' (%s).\n",
392			    progname, fn, strerror(errno));
393			return 1;
394		}
395
396		while ((len = getline(&line, &llen, f)) > 0) {
397			if (line[len - 1] == '\n')
398				line[len - 1] = '\0';
399			cwid = file_mbswidth(line);
400			if (cwid > wid)
401				wid = cwid;
402		}
403
404		rewind(f);
405	}
406
407	while ((len = getline(&line, &llen, f)) > 0) {
408		if (line[len - 1] == '\n')
409			line[len - 1] = '\0';
410		e |= process(ms, line, wid);
411		if(nobuffer)
412			(void)fflush(stdout);
413	}
414
415	free(line);
416	(void)fclose(f);
417	return e;
418}
419
420/*
421 * Called for each input file on the command line (or in a list of files)
422 */
423private int
424process(struct magic_set *ms, const char *inname, int wid)
425{
426	const char *type;
427	int std_in = strcmp(inname, "-") == 0;
428
429	if (wid > 0 && !bflag) {
430		(void)printf("%s", std_in ? "/dev/stdin" : inname);
431		if (nulsep)
432			(void)putc('\0', stdout);
433		(void)printf("%s", separator);
434		(void)printf("%*s ",
435		    (int) (nopad ? 0 : (wid - file_mbswidth(inname))), "");
436	}
437
438	type = magic_file(ms, std_in ? NULL : inname);
439	if (type == NULL) {
440		(void)printf("ERROR: %s\n", magic_error(ms));
441		return 1;
442	} else {
443		(void)printf("%s\n", type);
444		return 0;
445	}
446}
447
448protected size_t
449file_mbswidth(const char *s)
450{
451#if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH)
452	size_t bytesconsumed, old_n, n, width = 0;
453	mbstate_t state;
454	wchar_t nextchar;
455	(void)memset(&state, 0, sizeof(mbstate_t));
456	old_n = n = strlen(s);
457
458	while (n > 0) {
459		bytesconsumed = mbrtowc(&nextchar, s, n, &state);
460		if (bytesconsumed == (size_t)(-1) ||
461		    bytesconsumed == (size_t)(-2)) {
462			/* Something went wrong, return something reasonable */
463			return old_n;
464		}
465		if (s[0] == '\n') {
466			/*
467			 * do what strlen() would do, so that caller
468			 * is always right
469			 */
470			width++;
471		} else {
472			int w = wcwidth(nextchar);
473			if (w > 0)
474				width += w;
475		}
476
477		s += bytesconsumed, n -= bytesconsumed;
478	}
479	return width;
480#else
481	return strlen(s);
482#endif
483}
484
485private void
486usage(void)
487{
488	(void)fprintf(stderr, USAGE, progname, progname, progname);
489	exit(1);
490}
491
492private void
493docprint(const char *opts)
494{
495	size_t i;
496	int comma;
497	char *sp, *p;
498
499	p = strstr(opts, "%o");
500	if (p == NULL) {
501		fprintf(stdout, "%s", opts);
502		return;
503	}
504
505	for (sp = p - 1; sp > opts && *sp == ' '; sp--)
506		continue;
507
508	fprintf(stdout, "%.*s", (int)(p - opts), opts);
509
510	comma = 0;
511	for (i = 0; i < __arraycount(nv); i++) {
512		fprintf(stdout, "%s%s", comma++ ? ", " : "", nv[i].name);
513		if (i && i % 5 == 0) {
514			fprintf(stdout, ",\n%*s", (int)(p - sp - 1), "");
515			comma = 0;
516		}
517	}
518
519	fprintf(stdout, "%s", opts + (p - opts) + 2);
520}
521
522private void
523help(void)
524{
525	(void)fputs(
526"Usage: file [OPTION...] [FILE...]\n"
527"Determine type of FILEs.\n"
528"\n", stdout);
529#define OPT(shortname, longname, opt, doc)      \
530	fprintf(stdout, "  -%c, --" longname, shortname), \
531	docprint(doc);
532#define OPT_LONGONLY(longname, opt, doc)        \
533	fprintf(stdout, "      --" longname),	\
534	docprint(doc);
535#include "file_opts.h"
536#undef OPT
537#undef OPT_LONGONLY
538	fprintf(stdout, "\nReport bugs to http://bugs.gw.com/\n");
539	exit(0);
540}
541