magic.c revision 169942 
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("@(#)$Id: magic.c,v 1.35 2006/10/31 19:37:17 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 = malloc(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.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.off = malloc((ms->c.len = 10) * sizeof(*ms->c.off));
106	if (ms->c.off == NULL)
107		goto free3;
108
109	ms->o.len = 0;
110	ms->haderr = 0;
111	ms->error = -1;
112	ms->mlist = NULL;
113	return ms;
114free3:
115	free(ms->o.pbuf);
116free2:
117	free(ms->o.buf);
118free1:
119	free(ms);
120	return NULL;
121}
122
123private void
124free_mlist(struct mlist *mlist)
125{
126	struct mlist *ml;
127
128	if (mlist == NULL)
129		return;
130
131	for (ml = mlist->next; ml != mlist;) {
132		struct mlist *next = ml->next;
133		struct magic *mg = ml->magic;
134		file_delmagic(mg, ml->mapped, ml->nmagic);
135		free(ml);
136		ml = next;
137	}
138	free(ml);
139}
140
141private int
142info_from_stat(struct magic_set *ms, mode_t md)
143{
144	/* We cannot open it, but we were able to stat it. */
145	if (md & 0222)
146		if (file_printf(ms, "writable, ") == -1)
147			return -1;
148	if (md & 0111)
149		if (file_printf(ms, "executable, ") == -1)
150			return -1;
151	if (S_ISREG(md))
152		if (file_printf(ms, "regular file, ") == -1)
153			return -1;
154	if (file_printf(ms, "no read permission") == -1)
155		return -1;
156	return 0;
157}
158
159public void
160magic_close(struct magic_set *ms)
161{
162	free_mlist(ms->mlist);
163	free(ms->o.pbuf);
164	free(ms->o.buf);
165	free(ms->c.off);
166	free(ms);
167}
168
169/*
170 * load a magic file
171 */
172public int
173magic_load(struct magic_set *ms, const char *magicfile)
174{
175	struct mlist *ml = file_apprentice(ms, magicfile, FILE_LOAD);
176	if (ml) {
177		free_mlist(ms->mlist);
178		ms->mlist = ml;
179		return 0;
180	}
181	return -1;
182}
183
184public int
185magic_compile(struct magic_set *ms, const char *magicfile)
186{
187	struct mlist *ml = file_apprentice(ms, magicfile, FILE_COMPILE);
188	free_mlist(ml);
189	return ml ? 0 : -1;
190}
191
192public int
193magic_check(struct magic_set *ms, const char *magicfile)
194{
195	struct mlist *ml = file_apprentice(ms, magicfile, FILE_CHECK);
196	free_mlist(ml);
197	return ml ? 0 : -1;
198}
199
200private void
201close_and_restore(const struct magic_set *ms, const char *name, int fd,
202    const struct stat *sb)
203{
204	if (fd == STDIN_FILENO)
205		return;
206	(void) close(fd);
207
208	if ((ms->flags & MAGIC_PRESERVE_ATIME) != 0) {
209		/*
210		 * Try to restore access, modification times if read it.
211		 * This is really *bad* because it will modify the status
212		 * time of the file... And of course this will affect
213		 * backup programs
214		 */
215#ifdef HAVE_UTIMES
216		struct timeval  utsbuf[2];
217		utsbuf[0].tv_sec = sb->st_atime;
218		utsbuf[1].tv_sec = sb->st_mtime;
219
220		(void) utimes(name, utsbuf); /* don't care if loses */
221#elif defined(HAVE_UTIME_H) || defined(HAVE_SYS_UTIME_H)
222		struct utimbuf  utbuf;
223
224		utbuf.actime = sb->st_atime;
225		utbuf.modtime = sb->st_mtime;
226		(void) utime(name, &utbuf); /* don't care if loses */
227#endif
228	}
229}
230
231#ifndef COMPILE_ONLY
232/*
233 * find type of named file
234 */
235public const char *
236magic_file(struct magic_set *ms, const char *inname)
237{
238	int	fd = 0;
239	int	rv = -1;
240	unsigned char *buf;
241	struct stat	sb;
242	ssize_t nbytes = 0;	/* number of bytes read from a datafile */
243	int	ispipe = 0;
244
245	/*
246	 * one extra for terminating '\0', and
247	 * some overlapping space for matches near EOF
248	 */
249#define SLOP (1 + sizeof(union VALUETYPE))
250	if ((buf = malloc(HOWMANY + SLOP)) == NULL)
251		return NULL;
252
253	if (file_reset(ms) == -1)
254		goto done;
255
256	switch (file_fsmagic(ms, inname, &sb)) {
257	case -1:		/* error */
258		goto done;
259	case 0:			/* nothing found */
260		break;
261	default:		/* matched it and printed type */
262		rv = 0;
263		goto done;
264	}
265
266	if (inname == NULL) {
267		fd = STDIN_FILENO;
268		if (fstat(fd, &sb) == 0 && S_ISFIFO(sb.st_mode))
269			ispipe = 1;
270	} else {
271		int flags = O_RDONLY|O_BINARY;
272
273		if (stat(inname, &sb) == 0 && S_ISFIFO(sb.st_mode)) {
274			flags |= O_NONBLOCK;
275			ispipe = 1;
276		}
277
278		errno = 0;
279		if ((fd = open(inname, flags)) < 0) {
280#ifdef __CYGWIN__
281		    char *tmp = alloca(strlen(inname) + 5);
282		    (void)strcat(strcpy(tmp, inname), ".exe");
283		    if ((fd = open(tmp, flags)) < 0) {
284#endif
285			if (info_from_stat(ms, sb.st_mode) == -1)
286			    goto done;
287			rv = 0;
288			goto done;
289#ifdef __CYGWIN__
290		    }
291#endif
292		}
293#ifdef O_NONBLOCK
294		if ((flags = fcntl(fd, F_GETFL)) != -1) {
295			flags &= ~O_NONBLOCK;
296			(void)fcntl(fd, F_SETFL, flags);
297		}
298#endif
299	}
300
301	/*
302	 * try looking at the first HOWMANY bytes
303	 */
304	if (ispipe) {
305		ssize_t r = 0;
306
307		while ((r = sread(fd, (void *)&buf[nbytes],
308		    (size_t)(HOWMANY - nbytes))) > 0) {
309			nbytes += r;
310			if (r < PIPE_BUF) break;
311		}
312
313		if (nbytes == 0) {
314			/* We can not read it, but we were able to stat it. */
315			if (info_from_stat(ms, sb.st_mode) == -1)
316				goto done;
317			rv = 0;
318			goto done;
319		}
320
321	} else {
322		if ((nbytes = read(fd, (char *)buf, HOWMANY)) == -1) {
323			file_error(ms, errno, "cannot read `%s'", inname);
324			goto done;
325		}
326	}
327
328	if (nbytes == 0) {
329		if (file_printf(ms, (ms->flags & MAGIC_MIME) ?
330		    "application/x-empty" : "empty") == -1)
331			goto done;
332	} else if (nbytes == 1) {
333		if (file_printf(ms, "very short file (no magic)") == -1)
334			goto done;
335	} else {
336		(void)memset(buf + nbytes, 0, SLOP); /* NUL terminate */
337#ifdef __EMX__
338		switch (file_os2_apptype(ms, inname, buf, nbytes)) {
339		case -1:
340			goto done;
341		case 0:
342			break;
343		default:
344			rv = 0;
345			goto done;
346		}
347#endif
348		if (file_buffer(ms, fd, buf, (size_t)nbytes) == -1)
349			goto done;
350#ifdef BUILTIN_ELF
351		if (nbytes > 5) {
352			/*
353			 * We matched something in the file, so this *might*
354			 * be an ELF file, and the file is at least 5 bytes
355			 * long, so if it's an ELF file it has at least one
356			 * byte past the ELF magic number - try extracting
357			 * information from the ELF headers that cannot easily
358			 * be extracted with rules in the magic file.
359			 */
360			file_tryelf(ms, fd, buf, (size_t)nbytes);
361		}
362#endif
363	}
364	rv = 0;
365done:
366	free(buf);
367	close_and_restore(ms, inname, fd, &sb);
368	return rv == 0 ? file_getbuffer(ms) : NULL;
369}
370
371
372public const char *
373magic_buffer(struct magic_set *ms, const void *buf, size_t nb)
374{
375	if (file_reset(ms) == -1)
376		return NULL;
377	/*
378	 * The main work is done here!
379	 * We have the file name and/or the data buffer to be identified.
380	 */
381	if (file_buffer(ms, -1, buf, nb) == -1) {
382		return NULL;
383	}
384	return file_getbuffer(ms);
385}
386#endif
387
388public const char *
389magic_error(struct magic_set *ms)
390{
391	return ms->haderr ? ms->o.buf : NULL;
392}
393
394public int
395magic_errno(struct magic_set *ms)
396{
397	return ms->haderr ? ms->error : 0;
398}
399
400public int
401magic_setflags(struct magic_set *ms, int flags)
402{
403#if !defined(HAVE_UTIME) && !defined(HAVE_UTIMES)
404	if (flags & MAGIC_PRESERVE_ATIME)
405		return -1;
406#endif
407	ms->flags = flags;
408	return 0;
409}
410