magic.c revision 169962 
1/*
2 * Copyright (c) Christos Zoulas 2003.
3 * All Rights Reserved.
4 *
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions
7 * are met:
8 * 1. Redistributions of source code must retain the above copyright
9 *    notice immediately at the beginning of the file, without modification,
10 *    this list of conditions, and the following disclaimer.
11 * 2. Redistributions in binary form must reproduce the above copyright
12 *    notice, this list of conditions and the following disclaimer in the
13 *    documentation and/or other materials provided with the distribution.
14 *
15 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
16 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
17 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
18 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE FOR
19 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
20 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
21 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
22 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
23 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
24 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
25 * SUCH DAMAGE.
26 */
27
28#include "file.h"
29#include "magic.h"
30
31#include <stdio.h>
32#include <stdlib.h>
33#include <unistd.h>
34#include <string.h>
35#include <sys/types.h>
36#include <sys/param.h>	/* for MAXPATHLEN */
37#include <sys/stat.h>
38#ifdef QUICK
39#include <sys/mman.h>
40#endif
41#include <limits.h>	/* for PIPE_BUF */
42
43#if defined(HAVE_UTIMES)
44# include <sys/time.h>
45#elif defined(HAVE_UTIME)
46# if defined(HAVE_SYS_UTIME_H)
47#  include <sys/utime.h>
48# elif defined(HAVE_UTIME_H)
49#  include <utime.h>
50# endif
51#endif
52
53#ifdef HAVE_UNISTD_H
54#include <unistd.h>	/* for read() */
55#endif
56
57#ifdef HAVE_LOCALE_H
58#include <locale.h>
59#endif
60
61#include <netinet/in.h>		/* for byte swapping */
62
63#include "patchlevel.h"
64
65#ifndef	lint
66FILE_RCSID("@(#)$File: magic.c,v 1.41 2007/03/26 17:59:50 christos Exp $")
67#endif	/* lint */
68
69#ifdef __EMX__
70private char *apptypeName = NULL;
71protected int file_os2_apptype(struct magic_set *ms, const char *fn,
72    const void *buf, size_t nb);
73#endif /* __EMX__ */
74
75private void free_mlist(struct mlist *);
76private void close_and_restore(const struct magic_set *, const char *, int,
77    const struct stat *);
78private int info_from_stat(struct magic_set *, mode_t);
79
80#ifndef	STDIN_FILENO
81#define	STDIN_FILENO	0
82#endif
83
84public struct magic_set *
85magic_open(int flags)
86{
87	struct magic_set *ms;
88
89	if ((ms = calloc((size_t)1, sizeof(struct magic_set))) == NULL)
90		return NULL;
91
92	if (magic_setflags(ms, flags) == -1) {
93		errno = EINVAL;
94		goto free1;
95	}
96
97	ms->o.ptr = ms->o.buf = malloc(ms->o.left = ms->o.size = 1024);
98	if (ms->o.buf == NULL)
99		goto free1;
100
101	ms->o.pbuf = malloc(ms->o.psize = 1024);
102	if (ms->o.pbuf == NULL)
103		goto free2;
104
105	ms->c.li = malloc((ms->c.len = 10) * sizeof(*ms->c.li));
106	if (ms->c.li == NULL)
107		goto free3;
108
109	ms->haderr = 0;
110	ms->error = -1;
111	ms->mlist = NULL;
112	ms->file = "unknown";
113	ms->line = 0;
114	return ms;
115free3:
116	free(ms->o.pbuf);
117free2:
118	free(ms->o.buf);
119free1:
120	free(ms);
121	return NULL;
122}
123
124private void
125free_mlist(struct mlist *mlist)
126{
127	struct mlist *ml;
128
129	if (mlist == NULL)
130		return;
131
132	for (ml = mlist->next; ml != mlist;) {
133		struct mlist *next = ml->next;
134		struct magic *mg = ml->magic;
135		file_delmagic(mg, ml->mapped, ml->nmagic);
136		free(ml);
137		ml = next;
138	}
139	free(ml);
140}
141
142private int
143info_from_stat(struct magic_set *ms, mode_t md)
144{
145	/* We cannot open it, but we were able to stat it. */
146	if (md & 0222)
147		if (file_printf(ms, "writable, ") == -1)
148			return -1;
149	if (md & 0111)
150		if (file_printf(ms, "executable, ") == -1)
151			return -1;
152	if (S_ISREG(md))
153		if (file_printf(ms, "regular file, ") == -1)
154			return -1;
155	if (file_printf(ms, "no read permission") == -1)
156		return -1;
157	return 0;
158}
159
160public void
161magic_close(struct magic_set *ms)
162{
163	free_mlist(ms->mlist);
164	free(ms->o.pbuf);
165	free(ms->o.buf);
166	free(ms->c.li);
167	free(ms);
168}
169
170/*
171 * load a magic file
172 */
173public int
174magic_load(struct magic_set *ms, const char *magicfile)
175{
176	struct mlist *ml = file_apprentice(ms, magicfile, FILE_LOAD);
177	if (ml) {
178		free_mlist(ms->mlist);
179		ms->mlist = ml;
180		return 0;
181	}
182	return -1;
183}
184
185public int
186magic_compile(struct magic_set *ms, const char *magicfile)
187{
188	struct mlist *ml = file_apprentice(ms, magicfile, FILE_COMPILE);
189	free_mlist(ml);
190	return ml ? 0 : -1;
191}
192
193public int
194magic_check(struct magic_set *ms, const char *magicfile)
195{
196	struct mlist *ml = file_apprentice(ms, magicfile, FILE_CHECK);
197	free_mlist(ml);
198	return ml ? 0 : -1;
199}
200
201private void
202close_and_restore(const struct magic_set *ms, const char *name, int fd,
203    const struct stat *sb)
204{
205	if (fd == STDIN_FILENO)
206		return;
207	(void) close(fd);
208
209	if ((ms->flags & MAGIC_PRESERVE_ATIME) != 0) {
210		/*
211		 * Try to restore access, modification times if read it.
212		 * This is really *bad* because it will modify the status
213		 * time of the file... And of course this will affect
214		 * backup programs
215		 */
216#ifdef HAVE_UTIMES
217		struct timeval  utsbuf[2];
218		utsbuf[0].tv_sec = sb->st_atime;
219		utsbuf[1].tv_sec = sb->st_mtime;
220
221		(void) utimes(name, utsbuf); /* don't care if loses */
222#elif defined(HAVE_UTIME_H) || defined(HAVE_SYS_UTIME_H)
223		struct utimbuf  utbuf;
224
225		utbuf.actime = sb->st_atime;
226		utbuf.modtime = sb->st_mtime;
227		(void) utime(name, &utbuf); /* don't care if loses */
228#endif
229	}
230}
231
232#ifndef COMPILE_ONLY
233/*
234 * find type of named file
235 */
236public const char *
237magic_file(struct magic_set *ms, const char *inname)
238{
239	int	fd = 0;
240	int	rv = -1;
241	unsigned char *buf;
242	struct stat	sb;
243	ssize_t nbytes = 0;	/* number of bytes read from a datafile */
244	int	ispipe = 0;
245
246	/*
247	 * one extra for terminating '\0', and
248	 * some overlapping space for matches near EOF
249	 */
250#define SLOP (1 + sizeof(union VALUETYPE))
251	if ((buf = malloc(HOWMANY + SLOP)) == NULL)
252		return NULL;
253
254	if (file_reset(ms) == -1)
255		goto done;
256
257	switch (file_fsmagic(ms, inname, &sb)) {
258	case -1:		/* error */
259		goto done;
260	case 0:			/* nothing found */
261		break;
262	default:		/* matched it and printed type */
263		rv = 0;
264		goto done;
265	}
266
267	if (inname == NULL) {
268		fd = STDIN_FILENO;
269		if (fstat(fd, &sb) == 0 && S_ISFIFO(sb.st_mode))
270			ispipe = 1;
271	} else {
272		int flags = O_RDONLY|O_BINARY;
273
274		if (stat(inname, &sb) == 0 && S_ISFIFO(sb.st_mode)) {
275			flags |= O_NONBLOCK;
276			ispipe = 1;
277		}
278
279		errno = 0;
280		if ((fd = open(inname, flags)) < 0) {
281#ifdef __CYGWIN__
282		    char *tmp = alloca(strlen(inname) + 5);
283		    (void)strcat(strcpy(tmp, inname), ".exe");
284		    if ((fd = open(tmp, flags)) < 0) {
285#endif
286			if (info_from_stat(ms, sb.st_mode) == -1)
287			    goto done;
288			rv = 0;
289			goto done;
290#ifdef __CYGWIN__
291		    }
292#endif
293		}
294#ifdef O_NONBLOCK
295		if ((flags = fcntl(fd, F_GETFL)) != -1) {
296			flags &= ~O_NONBLOCK;
297			(void)fcntl(fd, F_SETFL, flags);
298		}
299#endif
300	}
301
302	/*
303	 * try looking at the first HOWMANY bytes
304	 */
305	if (ispipe) {
306		ssize_t r = 0;
307
308		while ((r = sread(fd, (void *)&buf[nbytes],
309		    (size_t)(HOWMANY - nbytes), 1)) > 0) {
310			nbytes += r;
311			if (r < PIPE_BUF) break;
312		}
313
314		if (nbytes == 0) {
315			/* We can not read it, but we were able to stat it. */
316			if (info_from_stat(ms, sb.st_mode) == -1)
317				goto done;
318			rv = 0;
319			goto done;
320		}
321
322	} else {
323		if ((nbytes = read(fd, (char *)buf, HOWMANY)) == -1) {
324			file_error(ms, errno, "cannot read `%s'", inname);
325			goto done;
326		}
327	}
328
329	if (nbytes == 0) {
330		if (file_printf(ms, (ms->flags & MAGIC_MIME) ?
331		    "application/x-empty" : "empty") == -1)
332			goto done;
333	} else if (nbytes == 1) {
334		if (file_printf(ms, "very short file (no magic)") == -1)
335			goto done;
336	} else {
337		(void)memset(buf + nbytes, 0, SLOP); /* NUL terminate */
338		if (file_buffer(ms, fd, inname, buf, (size_t)nbytes) == -1)
339			goto done;
340	}
341	rv = 0;
342done:
343	free(buf);
344	close_and_restore(ms, inname, fd, &sb);
345	return rv == 0 ? file_getbuffer(ms) : NULL;
346}
347
348
349public const char *
350magic_buffer(struct magic_set *ms, const void *buf, size_t nb)
351{
352	if (file_reset(ms) == -1)
353		return NULL;
354	/*
355	 * The main work is done here!
356	 * We have the file name and/or the data buffer to be identified.
357	 */
358	if (file_buffer(ms, -1, NULL, buf, nb) == -1) {
359		return NULL;
360	}
361	return file_getbuffer(ms);
362}
363#endif
364
365public const char *
366magic_error(struct magic_set *ms)
367{
368	return ms->haderr ? ms->o.buf : NULL;
369}
370
371public int
372magic_errno(struct magic_set *ms)
373{
374	return ms->haderr ? ms->error : 0;
375}
376
377public int
378magic_setflags(struct magic_set *ms, int flags)
379{
380#if !defined(HAVE_UTIME) && !defined(HAVE_UTIMES)
381	if (flags & MAGIC_PRESERVE_ATIME)
382		return -1;
383#endif
384	ms->flags = flags;
385	return 0;
386}
387