1/*	$NetBSD: file.c,v 1.1.1.2.6.1 2012/03/07 23:18:29 riz Exp $	*/
2
3/*
4 * Copyright (c) Ian F. Darwin 1986-1995.
5 * Software written by Ian F. Darwin and others;
6 * maintained 1995-present by Christos Zoulas and others.
7 *
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
10 * are met:
11 * 1. Redistributions of source code must retain the above copyright
12 *    notice immediately at the beginning of the file, without modification,
13 *    this list of conditions, and the following disclaimer.
14 * 2. Redistributions in binary form must reproduce the above copyright
15 *    notice, this list of conditions and the following disclaimer in the
16 *    documentation and/or other materials provided with the distribution.
17 *
18 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
19 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
20 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
21 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE FOR
22 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
23 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
24 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
25 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
26 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
27 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
28 * SUCH DAMAGE.
29 */
30/*
31 * file - find type of a file or files - main program.
32 */
33
34#include "file.h"
35
36#ifndef	lint
37#if 0
38FILE_RCSID("@(#)$File: file.c,v 1.145 2011/12/08 12:12:46 rrt Exp $")
39#else
40__RCSID("$NetBSD: file.c,v 1.1.1.2.6.1 2012/03/07 23:18:29 riz Exp $");
41#endif
42#endif	/* lint */
43
44#include "magic.h"
45
46#include <stdlib.h>
47#include <unistd.h>
48#include <string.h>
49#ifdef RESTORE_TIME
50# if (__COHERENT__ >= 0x420)
51#  include <sys/utime.h>
52# else
53#  ifdef USE_UTIMES
54#   include <sys/time.h>
55#  else
56#   include <utime.h>
57#  endif
58# endif
59#endif
60#ifdef HAVE_UNISTD_H
61#include <unistd.h>	/* for read() */
62#endif
63#ifdef HAVE_LOCALE_H
64#include <locale.h>
65#endif
66#ifdef HAVE_WCHAR_H
67#include <wchar.h>
68#endif
69
70#if defined(HAVE_GETOPT_H) && defined(HAVE_STRUCT_OPTION)
71#include <getopt.h>
72#ifndef HAVE_GETOPT_LONG
73int getopt_long(int argc, char * const *argv, const char *optstring, const struct option *longopts, int *longindex);
74#endif
75#else
76#include "mygetopt.h"
77#endif
78
79#ifdef S_IFLNK
80#define FILE_FLAGS "-bchikLlNnprsvz0"
81#else
82#define FILE_FLAGS "-bciklNnprsvz0"
83#endif
84
85# define USAGE  \
86    "Usage: %s [" FILE_FLAGS \
87	"] [--apple] [--mime-encoding] [--mime-type]\n" \
88    "            [-e testname] [-F separator] [-f namefile] [-m magicfiles] " \
89    "file ...\n" \
90    "       %s -C [-m magicfiles]\n" \
91    "       %s [--help]\n"
92
93private int 		/* Global command-line options 		*/
94	bflag = 0,	/* brief output format	 		*/
95	nopad = 0,	/* Don't pad output			*/
96	nobuffer = 0,   /* Do not buffer stdout 		*/
97	nulsep = 0;	/* Append '\0' to the separator		*/
98
99private const char *separator = ":";	/* Default field separator	*/
100private const struct option long_options[] = {
101#define OPT(shortname, longname, opt, doc)      \
102    {longname, opt, NULL, shortname},
103#define OPT_LONGONLY(longname, opt, doc)        \
104    {longname, opt, NULL, 0},
105#include "file_opts.h"
106#undef OPT
107#undef OPT_LONGONLY
108    {0, 0, NULL, 0}
109};
110#define OPTSTRING	"bcCde:f:F:hiklLm:nNprsvz0"
111
112private const struct {
113	const char *name;
114	int value;
115} nv[] = {
116	{ "apptype",	MAGIC_NO_CHECK_APPTYPE },
117	{ "ascii",	MAGIC_NO_CHECK_ASCII },
118	{ "cdf",	MAGIC_NO_CHECK_CDF },
119	{ "compress",	MAGIC_NO_CHECK_COMPRESS },
120	{ "elf",	MAGIC_NO_CHECK_ELF },
121	{ "encoding",	MAGIC_NO_CHECK_ENCODING },
122	{ "soft",	MAGIC_NO_CHECK_SOFT },
123	{ "tar",	MAGIC_NO_CHECK_TAR },
124	{ "text",	MAGIC_NO_CHECK_TEXT },	/* synonym for ascii */
125	{ "tokens",	MAGIC_NO_CHECK_TOKENS }, /* OBSOLETE: ignored for backwards compatibility */
126};
127
128private char *progname;		/* used throughout 		*/
129
130private void usage(void);
131private void help(void);
132int main(int, char *[]);
133
134private int unwrap(struct magic_set *, const char *);
135private int process(struct magic_set *ms, const char *, int);
136private struct magic_set *load(const char *, int);
137
138
139/*
140 * main - parse arguments and handle options
141 */
142int
143main(int argc, char *argv[])
144{
145	int c;
146	size_t i;
147	int action = 0, didsomefiles = 0, errflg = 0;
148	int flags = 0, e = 0;
149	struct magic_set *magic = NULL;
150	int longindex;
151	const char *magicfile = NULL;		/* where the magic is	*/
152
153	/* makes islower etc work for other langs */
154	(void)setlocale(LC_CTYPE, "");
155
156#ifdef __EMX__
157	/* sh-like wildcard expansion! Shouldn't hurt at least ... */
158	_wildcard(&argc, &argv);
159#endif
160
161	if ((progname = strrchr(argv[0], '/')) != NULL)
162		progname++;
163	else
164		progname = argv[0];
165
166#ifdef S_IFLNK
167	flags |= getenv("POSIXLY_CORRECT") ? MAGIC_SYMLINK : 0;
168#endif
169	while ((c = getopt_long(argc, argv, OPTSTRING, long_options,
170	    &longindex)) != -1)
171		switch (c) {
172		case 0 :
173			switch (longindex) {
174			case 0:
175				help();
176				break;
177			case 10:
178				flags |= MAGIC_APPLE;
179				break;
180			case 11:
181				flags |= MAGIC_MIME_TYPE;
182				break;
183			case 12:
184				flags |= MAGIC_MIME_ENCODING;
185				break;
186			}
187			break;
188		case '0':
189			nulsep = 1;
190			break;
191		case 'b':
192			bflag++;
193			break;
194		case 'c':
195			action = FILE_CHECK;
196			break;
197		case 'C':
198			action = FILE_COMPILE;
199			break;
200		case 'd':
201			flags |= MAGIC_DEBUG|MAGIC_CHECK;
202			break;
203		case 'e':
204			for (i = 0; i < sizeof(nv) / sizeof(nv[0]); i++)
205				if (strcmp(nv[i].name, optarg) == 0)
206					break;
207
208			if (i == sizeof(nv) / sizeof(nv[0]))
209				errflg++;
210			else
211				flags |= nv[i].value;
212			break;
213
214		case 'f':
215			if(action)
216				usage();
217			if (magic == NULL)
218				if ((magic = load(magicfile, flags)) == NULL)
219					return 1;
220			e |= unwrap(magic, optarg);
221			++didsomefiles;
222			break;
223		case 'F':
224			separator = optarg;
225			break;
226		case 'i':
227			flags |= MAGIC_MIME;
228			break;
229		case 'k':
230			flags |= MAGIC_CONTINUE;
231			break;
232		case 'l':
233			action = FILE_LIST;
234			break;
235		case 'm':
236			magicfile = optarg;
237			break;
238		case 'n':
239			++nobuffer;
240			break;
241		case 'N':
242			++nopad;
243			break;
244#if defined(HAVE_UTIME) || defined(HAVE_UTIMES)
245		case 'p':
246			flags |= MAGIC_PRESERVE_ATIME;
247			break;
248#endif
249		case 'r':
250			flags |= MAGIC_RAW;
251			break;
252		case 's':
253			flags |= MAGIC_DEVICES;
254			break;
255		case 'v':
256			if (magicfile == NULL)
257				magicfile = magic_getpath(magicfile, action);
258			(void)fprintf(stdout, "%s-%s\n", progname, VERSION);
259			(void)fprintf(stdout, "magic file from %s\n",
260				       magicfile);
261			return 1;
262		case 'z':
263			flags |= MAGIC_COMPRESS;
264			break;
265#ifdef S_IFLNK
266		case 'L':
267			flags |= MAGIC_SYMLINK;
268			break;
269		case 'h':
270			flags &= ~MAGIC_SYMLINK;
271			break;
272#endif
273		case '?':
274		default:
275			errflg++;
276			break;
277		}
278
279	if (errflg) {
280		usage();
281	}
282	if (e)
283		return e;
284
285	switch(action) {
286	case FILE_CHECK:
287	case FILE_COMPILE:
288	case FILE_LIST:
289		/*
290		 * Don't try to check/compile ~/.magic unless we explicitly
291		 * ask for it.
292		 */
293		magic = magic_open(flags|MAGIC_CHECK);
294		if (magic == NULL) {
295			(void)fprintf(stderr, "%s: %s\n", progname,
296			    strerror(errno));
297			return 1;
298		}
299		switch(action) {
300		case FILE_CHECK:
301			c = magic_check(magic, magicfile);
302			break;
303		case FILE_COMPILE:
304			c = magic_compile(magic, magicfile);
305			break;
306		case FILE_LIST:
307			c = magic_list(magic, magicfile);
308			break;
309		default:
310			abort();
311		}
312		if (c == -1) {
313			(void)fprintf(stderr, "%s: %s\n", progname,
314			    magic_error(magic));
315			return 1;
316		}
317		return 0;
318	default:
319		if (magic == NULL)
320			if ((magic = load(magicfile, flags)) == NULL)
321				return 1;
322		break;
323	}
324
325	if (optind == argc) {
326		if (!didsomefiles)
327			usage();
328	}
329	else {
330		size_t j, wid, nw;
331		for (wid = 0, j = (size_t)optind; j < (size_t)argc; j++) {
332			nw = file_mbswidth(argv[j]);
333			if (nw > wid)
334				wid = nw;
335		}
336		/*
337		 * If bflag is only set twice, set it depending on
338		 * number of files [this is undocumented, and subject to change]
339		 */
340		if (bflag == 2) {
341			bflag = optind >= argc - 1;
342		}
343		for (; optind < argc; optind++)
344			e |= process(magic, argv[optind], wid);
345	}
346
347	if (magic)
348		magic_close(magic);
349	return e;
350}
351
352
353private struct magic_set *
354/*ARGSUSED*/
355load(const char *magicfile, int flags)
356{
357	struct magic_set *magic = magic_open(flags);
358	if (magic == NULL) {
359		(void)fprintf(stderr, "%s: %s\n", progname, strerror(errno));
360		return NULL;
361	}
362	if (magic_load(magic, magicfile) == -1) {
363		(void)fprintf(stderr, "%s: %s\n",
364		    progname, magic_error(magic));
365		magic_close(magic);
366		return NULL;
367	}
368	return magic;
369}
370
371/*
372 * unwrap -- read a file of filenames, do each one.
373 */
374private int
375unwrap(struct magic_set *ms, const char *fn)
376{
377	FILE *f;
378	ssize_t len;
379	char *line = NULL;
380	size_t llen = 0;
381	int wid = 0, cwid;
382	int e = 0;
383
384	if (strcmp("-", fn) == 0) {
385		f = stdin;
386		wid = 1;
387	} else {
388		if ((f = fopen(fn, "r")) == NULL) {
389			(void)fprintf(stderr, "%s: Cannot open `%s' (%s).\n",
390			    progname, fn, strerror(errno));
391			return 1;
392		}
393
394		while ((len = getline(&line, &llen, f)) > 0) {
395			if (line[len - 1] == '\n')
396				line[len - 1] = '\0';
397			cwid = file_mbswidth(line);
398			if (cwid > wid)
399				wid = cwid;
400		}
401
402		rewind(f);
403	}
404
405	while ((len = getline(&line, &llen, f)) > 0) {
406		if (line[len - 1] == '\n')
407			line[len - 1] = '\0';
408		e |= process(ms, line, wid);
409		if(nobuffer)
410			(void)fflush(stdout);
411	}
412
413	free(line);
414	(void)fclose(f);
415	return e;
416}
417
418/*
419 * Called for each input file on the command line (or in a list of files)
420 */
421private int
422process(struct magic_set *ms, const char *inname, int wid)
423{
424	const char *type;
425	int std_in = strcmp(inname, "-") == 0;
426
427	if (wid > 0 && !bflag) {
428		(void)printf("%s", std_in ? "/dev/stdin" : inname);
429		if (nulsep)
430			(void)putc('\0', stdout);
431		(void)printf("%s", separator);
432		(void)printf("%*s ",
433		    (int) (nopad ? 0 : (wid - file_mbswidth(inname))), "");
434	}
435
436	type = magic_file(ms, std_in ? NULL : inname);
437	if (type == NULL) {
438		(void)printf("ERROR: %s\n", magic_error(ms));
439		return 1;
440	} else {
441		(void)printf("%s\n", type);
442		return 0;
443	}
444}
445
446size_t
447file_mbswidth(const char *s)
448{
449#if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH)
450	size_t bytesconsumed, old_n, n, width = 0;
451	mbstate_t state;
452	wchar_t nextchar;
453	(void)memset(&state, 0, sizeof(mbstate_t));
454	old_n = n = strlen(s);
455
456	while (n > 0) {
457		bytesconsumed = mbrtowc(&nextchar, s, n, &state);
458		if (bytesconsumed == (size_t)(-1) ||
459		    bytesconsumed == (size_t)(-2)) {
460			/* Something went wrong, return something reasonable */
461			return old_n;
462		}
463		if (s[0] == '\n') {
464			/*
465			 * do what strlen() would do, so that caller
466			 * is always right
467			 */
468			width++;
469		} else
470			width += wcwidth(nextchar);
471
472		s += bytesconsumed, n -= bytesconsumed;
473	}
474	return width;
475#else
476	return strlen(s);
477#endif
478}
479
480private void
481usage(void)
482{
483	(void)fprintf(stderr, USAGE, progname, progname, progname);
484	exit(1);
485}
486
487private void
488help(void)
489{
490	(void)fputs(
491"Usage: file [OPTION...] [FILE...]\n"
492"Determine type of FILEs.\n"
493"\n", stdout);
494#define OPT(shortname, longname, opt, doc)      \
495	fprintf(stdout, "  -%c, --" longname doc, shortname);
496#define OPT_LONGONLY(longname, opt, doc)        \
497	fprintf(stdout, "      --" longname doc);
498#include "file_opts.h"
499#undef OPT
500#undef OPT_LONGONLY
501	fprintf(stdout, "\nReport bugs to http://bugs.gw.com/\n");
502	exit(0);
503}
504