magic.c revision 1.1.1.2 
1/*	$NetBSD: magic.c,v 1.1.1.2 2011/05/12 20:46:50 christos Exp $	*/
2
3/*
4 * Copyright (c) Christos Zoulas 2003.
5 * All Rights Reserved.
6 *
7 * Redistribution and use in source and binary forms, with or without
8 * modification, are permitted provided that the following conditions
9 * are met:
10 * 1. Redistributions of source code must retain the above copyright
11 *    notice immediately at the beginning of the file, without modification,
12 *    this list of conditions, and the following disclaimer.
13 * 2. Redistributions in binary form must reproduce the above copyright
14 *    notice, this list of conditions and the following disclaimer in the
15 *    documentation and/or other materials provided with the distribution.
16 *
17 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
18 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE FOR
21 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
27 * SUCH DAMAGE.
28 */
29
30#ifdef WIN32
31#include <windows.h>
32#include <shlwapi.h>
33#endif
34
35#include "file.h"
36
37#ifndef	lint
38#if 0
39FILE_RCSID("@(#)$File: magic.c,v 1.73 2011/05/10 17:08:14 christos Exp $")
40#else
41__RCSID("$NetBSD: magic.c,v 1.1.1.2 2011/05/12 20:46:50 christos Exp $");
42#endif
43#endif	/* lint */
44
45#include "magic.h"
46
47#include <stdlib.h>
48#include <unistd.h>
49#include <string.h>
50#ifdef QUICK
51#include <sys/mman.h>
52#endif
53#ifdef HAVE_LIMITS_H
54#include <limits.h>	/* for PIPE_BUF */
55#endif
56
57#if defined(HAVE_UTIMES)
58# include <sys/time.h>
59#elif defined(HAVE_UTIME)
60# if defined(HAVE_SYS_UTIME_H)
61#  include <sys/utime.h>
62# elif defined(HAVE_UTIME_H)
63#  include <utime.h>
64# endif
65#endif
66
67#ifdef HAVE_UNISTD_H
68#include <unistd.h>	/* for read() */
69#endif
70
71#ifndef PIPE_BUF
72/* Get the PIPE_BUF from pathconf */
73#ifdef _PC_PIPE_BUF
74#define PIPE_BUF pathconf(".", _PC_PIPE_BUF)
75#else
76#define PIPE_BUF 512
77#endif
78#endif
79
80private void free_mlist(struct mlist *);
81private void close_and_restore(const struct magic_set *, const char *, int,
82    const struct stat *);
83private int unreadable_info(struct magic_set *, mode_t, const char *);
84private const char* get_default_magic(void);
85#ifndef COMPILE_ONLY
86private const char *file_or_fd(struct magic_set *, const char *, int);
87#endif
88
89#ifndef	STDIN_FILENO
90#define	STDIN_FILENO	0
91#endif
92
93#ifdef WIN32
94BOOL WINAPI DllMain(HINSTANCE hinstDLL,
95    DWORD fdwReason __attribute__((__unused__)),
96    LPVOID lpvReserved __attribute__((__unused__)));
97
98CHAR dllpath[MAX_PATH + 1] = { 0 };
99
100BOOL WINAPI DllMain(HINSTANCE hinstDLL,
101    DWORD fdwReason __attribute__((__unused__)),
102    LPVOID lpvReserved __attribute__((__unused__)))
103{
104	if (dllpath[0] == 0 &&
105	    GetModuleFileNameA(hinstDLL, dllpath, MAX_PATH) != 0)
106		PathRemoveFileSpecA(dllpath);
107	return TRUE;
108}
109#endif
110
111private const char *
112get_default_magic(void)
113{
114	static const char hmagic[] = "/.magic/magic.mgc";
115	static char *default_magic;
116	char *home, *hmagicpath;
117
118#ifndef WIN32
119	struct stat st;
120
121	if (default_magic) {
122		free(default_magic);
123		default_magic = NULL;
124	}
125	if ((home = getenv("HOME")) == NULL)
126		return MAGIC;
127
128	if (asprintf(&hmagicpath, "%s/.magic", home) < 0)
129		return MAGIC;
130	if (stat(hmagicpath, &st) == -1)
131		goto out;
132	if (S_ISDIR(st.st_mode)) {
133		free(hmagicpath);
134		if (asprintf(&hmagicpath, "%s/%s", home, hmagic) < 0)
135			return MAGIC;
136		if (access(hmagicpath, R_OK) == -1)
137			goto out;
138	}
139
140	if (asprintf(&default_magic, "%s:%s", hmagicpath, MAGIC) < 0)
141		goto out;
142	free(hmagicpath);
143	return default_magic;
144out:
145	default_magic = NULL;
146	free(hmagicpath);
147	return MAGIC;
148#else
149	char *hmagicp = hmagicpath;
150	char *tmppath = NULL;
151
152#define APPENDPATH() \
153	do { \
154		if (tmppath && access(tmppath, R_OK) != -1) { \
155			if (hmagicpath == NULL) { \
156				hmagicpath = tmppath; \
157				tmppath = NULL; \
158			} else { \
159				free(tmppath); \
160				if (asprintf(&hmagicp, "%s%c%s", hmagicpath, \
161				    PATHSEP, tmppath) >= 0) { \
162					free(hmagicpath); \
163					hmagicpath = hmagicp; \
164				} \
165			} \
166	} while (/*CONSTCOND*/0)
167
168	if (default_magic) {
169		free(default_magic);
170		default_magic = NULL;
171	}
172
173	/* First, try to get user-specific magic file */
174	if ((home = getenv("LOCALAPPDATA")) == NULL) {
175		if ((home = getenv("USERPROFILE")) != NULL)
176			if (asprintf(&tmppath,
177			    "%s/Local Settings/Application Data%s", home,
178			    hmagic) < 0)
179				tmppath = NULL;
180	} else {
181		if (asprintf(&tmppath, "%s%s", home, hmagic) < 0)
182			tmppath = NULL;
183	}
184
185	APPENDPATH();
186
187	/* Second, try to get a magic file from Common Files */
188	if ((home = getenv("COMMONPROGRAMFILES")) != NULL) {
189		if (asprintf(&tmppath, "%s%s", home, hmagic) >= 0)
190			APPENDPATH();
191	}
192
193
194	/* Third, try to get magic file relative to dll location */
195	if (dllpath[0] != 0) {
196		if (strlen(dllpath) > 3 &&
197		    stricmp(&dllpath[strlen(dllpath) - 3], "bin") == 0) {
198			if (asprintf(&tmppath,
199			    "%s/../share/misc/magic.mgc", dllpath) >= 0)
200				APPENDPATH();
201		} else {
202			if (asprintf(&tmppath,
203			    "%s/share/misc/magic.mgc", dllpath) >= 0)
204				APPENDPATH();
205			else if (asprintf(&tmppath,
206			    "%s/magic.mgc", dllpath) >= 0)
207				APPENDPATH();
208		}
209	}
210
211	/* Don't put MAGIC constant - it likely points to a file within MSys
212	tree */
213	default_magic = hmagicpath;
214	return default_magic;
215#endif
216}
217
218public const char *
219magic_getpath(const char *magicfile, int action)
220{
221	if (magicfile != NULL)
222		return magicfile;
223
224	magicfile = getenv("MAGIC");
225	if (magicfile != NULL)
226		return magicfile;
227
228	return action == FILE_LOAD ? get_default_magic() : MAGIC;
229}
230
231public struct magic_set *
232magic_open(int flags)
233{
234	struct magic_set *ms;
235	size_t len;
236
237	if ((ms = CAST(struct magic_set *, calloc((size_t)1,
238	    sizeof(struct magic_set)))) == NULL)
239		return NULL;
240
241	if (magic_setflags(ms, flags) == -1) {
242		errno = EINVAL;
243		goto free;
244	}
245
246	ms->o.buf = ms->o.pbuf = NULL;
247	len = (ms->c.len = 10) * sizeof(*ms->c.li);
248
249	if ((ms->c.li = CAST(struct level_info *, malloc(len))) == NULL)
250		goto free;
251
252	ms->event_flags = 0;
253	ms->error = -1;
254	ms->mlist = NULL;
255	ms->file = "unknown";
256	ms->line = 0;
257	return ms;
258free:
259	free(ms);
260	return NULL;
261}
262
263private void
264free_mlist(struct mlist *mlist)
265{
266	struct mlist *ml;
267
268	if (mlist == NULL)
269		return;
270
271	for (ml = mlist->next; ml != mlist;) {
272		struct mlist *next = ml->next;
273		struct magic *mg = ml->magic;
274		file_delmagic(mg, ml->mapped, ml->nmagic);
275		free(ml);
276		ml = next;
277	}
278	free(ml);
279}
280
281private int
282unreadable_info(struct magic_set *ms, mode_t md, const char *file)
283{
284	/* We cannot open it, but we were able to stat it. */
285	if (access(file, W_OK) == 0)
286		if (file_printf(ms, "writable, ") == -1)
287			return -1;
288	if (access(file, X_OK) == 0)
289		if (file_printf(ms, "executable, ") == -1)
290			return -1;
291	if (S_ISREG(md))
292		if (file_printf(ms, "regular file, ") == -1)
293			return -1;
294	if (file_printf(ms, "no read permission") == -1)
295		return -1;
296	return 0;
297}
298
299public void
300magic_close(struct magic_set *ms)
301{
302	free_mlist(ms->mlist);
303	free(ms->o.pbuf);
304	free(ms->o.buf);
305	free(ms->c.li);
306	free(ms);
307}
308
309/*
310 * load a magic file
311 */
312public int
313magic_load(struct magic_set *ms, const char *magicfile)
314{
315	struct mlist *ml = file_apprentice(ms, magicfile, FILE_LOAD);
316	if (ml) {
317		free_mlist(ms->mlist);
318		ms->mlist = ml;
319		return 0;
320	}
321	return -1;
322}
323
324public int
325magic_compile(struct magic_set *ms, const char *magicfile)
326{
327	struct mlist *ml = file_apprentice(ms, magicfile, FILE_COMPILE);
328	free_mlist(ml);
329	return ml ? 0 : -1;
330}
331
332public int
333magic_check(struct magic_set *ms, const char *magicfile)
334{
335	struct mlist *ml = file_apprentice(ms, magicfile, FILE_CHECK);
336	free_mlist(ml);
337	return ml ? 0 : -1;
338}
339
340public int
341magic_list(struct magic_set *ms, const char *magicfile)
342{
343	struct mlist *ml = file_apprentice(ms, magicfile, FILE_LIST);
344	free_mlist(ml);
345	return ml ? 0 : -1;
346}
347
348private void
349close_and_restore(const struct magic_set *ms, const char *name, int fd,
350    const struct stat *sb)
351{
352	if (fd == STDIN_FILENO)
353		return;
354	(void) close(fd);
355
356	if ((ms->flags & MAGIC_PRESERVE_ATIME) != 0) {
357		/*
358		 * Try to restore access, modification times if read it.
359		 * This is really *bad* because it will modify the status
360		 * time of the file... And of course this will affect
361		 * backup programs
362		 */
363#ifdef HAVE_UTIMES
364		struct timeval  utsbuf[2];
365		(void)memset(utsbuf, 0, sizeof(utsbuf));
366		utsbuf[0].tv_sec = sb->st_atime;
367		utsbuf[1].tv_sec = sb->st_mtime;
368
369		(void) utimes(name, utsbuf); /* don't care if loses */
370#elif defined(HAVE_UTIME_H) || defined(HAVE_SYS_UTIME_H)
371		struct utimbuf  utbuf;
372
373		(void)memset(&utbuf, 0, sizeof(utbuf));
374		utbuf.actime = sb->st_atime;
375		utbuf.modtime = sb->st_mtime;
376		(void) utime(name, &utbuf); /* don't care if loses */
377#endif
378	}
379}
380
381#ifndef COMPILE_ONLY
382
383/*
384 * find type of descriptor
385 */
386public const char *
387magic_descriptor(struct magic_set *ms, int fd)
388{
389	return file_or_fd(ms, NULL, fd);
390}
391
392/*
393 * find type of named file
394 */
395public const char *
396magic_file(struct magic_set *ms, const char *inname)
397{
398	return file_or_fd(ms, inname, STDIN_FILENO);
399}
400
401private const char *
402file_or_fd(struct magic_set *ms, const char *inname, int fd)
403{
404	int	rv = -1;
405	unsigned char *buf;
406	struct stat	sb;
407	ssize_t nbytes = 0;	/* number of bytes read from a datafile */
408	int	ispipe = 0;
409
410	/*
411	 * one extra for terminating '\0', and
412	 * some overlapping space for matches near EOF
413	 */
414#define SLOP (1 + sizeof(union VALUETYPE))
415	if ((buf = CAST(unsigned char *, malloc(HOWMANY + SLOP))) == NULL)
416		return NULL;
417
418	if (file_reset(ms) == -1)
419		goto done;
420
421	switch (file_fsmagic(ms, inname, &sb)) {
422	case -1:		/* error */
423		goto done;
424	case 0:			/* nothing found */
425		break;
426	default:		/* matched it and printed type */
427		rv = 0;
428		goto done;
429	}
430
431	if (inname == NULL) {
432		if (fstat(fd, &sb) == 0 && S_ISFIFO(sb.st_mode))
433			ispipe = 1;
434	} else {
435		int flags = O_RDONLY|O_BINARY;
436
437		if (stat(inname, &sb) == 0 && S_ISFIFO(sb.st_mode)) {
438#ifdef O_NONBLOCK
439			flags |= O_NONBLOCK;
440#endif
441			ispipe = 1;
442		}
443
444		errno = 0;
445		if ((fd = open(inname, flags)) < 0) {
446			if (unreadable_info(ms, sb.st_mode, inname) == -1)
447				goto done;
448			rv = 0;
449			goto done;
450		}
451#ifdef O_NONBLOCK
452		if ((flags = fcntl(fd, F_GETFL)) != -1) {
453			flags &= ~O_NONBLOCK;
454			(void)fcntl(fd, F_SETFL, flags);
455		}
456#endif
457	}
458
459	/*
460	 * try looking at the first HOWMANY bytes
461	 */
462	if (ispipe) {
463		ssize_t r = 0;
464
465		while ((r = sread(fd, (void *)&buf[nbytes],
466		    (size_t)(HOWMANY - nbytes), 1)) > 0) {
467			nbytes += r;
468			if (r < PIPE_BUF) break;
469		}
470
471		if (nbytes == 0) {
472			/* We can not read it, but we were able to stat it. */
473			if (unreadable_info(ms, sb.st_mode, inname) == -1)
474				goto done;
475			rv = 0;
476			goto done;
477		}
478
479	} else {
480		if ((nbytes = read(fd, (char *)buf, HOWMANY)) == -1) {
481			file_error(ms, errno, "cannot read `%s'", inname);
482			goto done;
483		}
484	}
485
486	(void)memset(buf + nbytes, 0, SLOP); /* NUL terminate */
487	if (file_buffer(ms, fd, inname, buf, (size_t)nbytes) == -1)
488		goto done;
489	rv = 0;
490done:
491	free(buf);
492	close_and_restore(ms, inname, fd, &sb);
493	return rv == 0 ? file_getbuffer(ms) : NULL;
494}
495
496
497public const char *
498magic_buffer(struct magic_set *ms, const void *buf, size_t nb)
499{
500	if (file_reset(ms) == -1)
501		return NULL;
502	/*
503	 * The main work is done here!
504	 * We have the file name and/or the data buffer to be identified.
505	 */
506	if (file_buffer(ms, -1, NULL, buf, nb) == -1) {
507		return NULL;
508	}
509	return file_getbuffer(ms);
510}
511#endif
512
513public const char *
514magic_error(struct magic_set *ms)
515{
516	return (ms->event_flags & EVENT_HAD_ERR) ? ms->o.buf : NULL;
517}
518
519public int
520magic_errno(struct magic_set *ms)
521{
522	return (ms->event_flags & EVENT_HAD_ERR) ? ms->error : 0;
523}
524
525public int
526magic_setflags(struct magic_set *ms, int flags)
527{
528#if !defined(HAVE_UTIME) && !defined(HAVE_UTIMES)
529	if (flags & MAGIC_PRESERVE_ATIME)
530		return -1;
531#endif
532	ms->flags = flags;
533	return 0;
534}
535