compress.c revision 299236 
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 * compress routines:
30 *	zmagic() - returns 0 if not recognized, uncompresses and prints
31 *		   information if recognized
32 *	uncompress(method, old, n, newch) - uncompress old into new,
33 *					    using method, return sizeof new
34 */
35#include "file.h"
36
37#ifndef lint
38FILE_RCSID("@(#)$File: compress.c,v 1.80 2015/06/03 18:21:24 christos Exp $")
39#endif
40
41#include "magic.h"
42#include <stdlib.h>
43#ifdef HAVE_UNISTD_H
44#include <unistd.h>
45#endif
46#include <string.h>
47#include <errno.h>
48#ifdef HAVE_SIGNAL_H
49#include <signal.h>
50# ifndef HAVE_SIG_T
51typedef void (*sig_t)(int);
52# endif /* HAVE_SIG_T */
53#endif
54#if !defined(__MINGW32__) && !defined(WIN32)
55#include <sys/ioctl.h>
56#endif
57#ifdef HAVE_SYS_WAIT_H
58#include <sys/wait.h>
59#endif
60#if defined(HAVE_SYS_TIME_H)
61#include <sys/time.h>
62#endif
63#if defined(HAVE_ZLIB_H) && defined(HAVE_LIBZ)
64#define BUILTIN_DECOMPRESS
65#include <zlib.h>
66#endif
67
68private const struct {
69	const char magic[8];
70	size_t maglen;
71	const char *argv[3];
72	int silent;
73} compr[] = {
74	{ "\037\235", 2, { "gzip", "-cdq", NULL }, 1 },		/* compressed */
75	/* Uncompress can get stuck; so use gzip first if we have it
76	 * Idea from Damien Clark, thanks! */
77	{ "\037\235", 2, { "uncompress", "-c", NULL }, 1 },	/* compressed */
78	{ "\037\213", 2, { "gzip", "-cdq", NULL }, 1 },		/* gzipped */
79	{ "\037\236", 2, { "gzip", "-cdq", NULL }, 1 },		/* frozen */
80	{ "\037\240", 2, { "gzip", "-cdq", NULL }, 1 },		/* SCO LZH */
81	/* the standard pack utilities do not accept standard input */
82	{ "\037\036", 2, { "gzip", "-cdq", NULL }, 0 },		/* packed */
83	{ "PK\3\4",   4, { "gzip", "-cdq", NULL }, 1 },		/* pkzipped, */
84					    /* ...only first file examined */
85	{ "BZh",      3, { "bzip2", "-cd", NULL }, 1 },		/* bzip2-ed */
86	{ "LZIP",     4, { "lzip", "-cdq", NULL }, 1 },
87 	{ "\3757zXZ\0",6,{ "xz", "-cd", NULL }, 1 },		/* XZ Utils */
88 	{ "LRZI",     4, { "lrzip", "-dqo-", NULL }, 1 },	/* LRZIP */
89 	{ "\004\"M\030", 4, { "lz4", "-cd", NULL }, 1 },	/* LZ4 */
90};
91
92#define NODATA ((size_t)~0)
93
94private ssize_t swrite(int, const void *, size_t);
95#if HAVE_FORK
96private size_t ncompr = sizeof(compr) / sizeof(compr[0]);
97private size_t uncompressbuf(struct magic_set *, int, size_t,
98    const unsigned char *, unsigned char **, size_t);
99#ifdef BUILTIN_DECOMPRESS
100private size_t uncompressgzipped(struct magic_set *, const unsigned char *,
101    unsigned char **, size_t);
102#endif
103
104protected int
105file_zmagic(struct magic_set *ms, int fd, const char *name,
106    const unsigned char *buf, size_t nbytes)
107{
108	unsigned char *newbuf = NULL;
109	size_t i, nsz;
110	int rv = 0;
111	int mime = ms->flags & MAGIC_MIME;
112#ifdef HAVE_SIGNAL_H
113	sig_t osigpipe;
114#endif
115
116	if ((ms->flags & MAGIC_COMPRESS) == 0)
117		return 0;
118
119#ifdef HAVE_SIGNAL_H
120	osigpipe = signal(SIGPIPE, SIG_IGN);
121#endif
122	for (i = 0; i < ncompr; i++) {
123		if (nbytes < compr[i].maglen)
124			continue;
125		if (memcmp(buf, compr[i].magic, compr[i].maglen) == 0 &&
126		    (nsz = uncompressbuf(ms, fd, i, buf, &newbuf,
127		    nbytes)) != NODATA) {
128			ms->flags &= ~MAGIC_COMPRESS;
129			rv = -1;
130			if (file_buffer(ms, -1, name, newbuf, nsz) == -1)
131				goto error;
132
133			if ((ms->flags & MAGIC_COMPRESS_TRANSP) == 0 &&
134			    (mime == MAGIC_MIME || mime == 0)) {
135				if (file_printf(ms, mime ?
136				    " compressed-encoding=" : " (") == -1)
137					goto error;
138				if (file_buffer(ms, -1, NULL, buf, nbytes) == -1)
139					goto error;
140				if (!mime && file_printf(ms, ")") == -1)
141					goto error;
142			}
143
144			rv = 1;
145			break;
146		}
147	}
148error:
149#ifdef HAVE_SIGNAL_H
150	(void)signal(SIGPIPE, osigpipe);
151#endif
152	free(newbuf);
153	ms->flags |= MAGIC_COMPRESS;
154	return rv;
155}
156#endif
157/*
158 * `safe' write for sockets and pipes.
159 */
160private ssize_t
161swrite(int fd, const void *buf, size_t n)
162{
163	ssize_t rv;
164	size_t rn = n;
165
166	do
167		switch (rv = write(fd, buf, n)) {
168		case -1:
169			if (errno == EINTR)
170				continue;
171			return -1;
172		default:
173			n -= rv;
174			buf = CAST(const char *, buf) + rv;
175			break;
176		}
177	while (n > 0);
178	return rn;
179}
180
181
182/*
183 * `safe' read for sockets and pipes.
184 */
185protected ssize_t
186sread(int fd, void *buf, size_t n, int canbepipe __attribute__((__unused__)))
187{
188	ssize_t rv;
189#ifdef FIONREAD
190	int t = 0;
191#endif
192	size_t rn = n;
193
194	if (fd == STDIN_FILENO)
195		goto nocheck;
196
197#ifdef FIONREAD
198	if (canbepipe && (ioctl(fd, FIONREAD, &t) == -1 || t == 0)) {
199#ifdef FD_ZERO
200		ssize_t cnt;
201		for (cnt = 0;; cnt++) {
202			fd_set check;
203			struct timeval tout = {0, 100 * 1000};
204			int selrv;
205
206			FD_ZERO(&check);
207			FD_SET(fd, &check);
208
209			/*
210			 * Avoid soft deadlock: do not read if there
211			 * is nothing to read from sockets and pipes.
212			 */
213			selrv = select(fd + 1, &check, NULL, NULL, &tout);
214			if (selrv == -1) {
215				if (errno == EINTR || errno == EAGAIN)
216					continue;
217			} else if (selrv == 0 && cnt >= 5) {
218				return 0;
219			} else
220				break;
221		}
222#endif
223		(void)ioctl(fd, FIONREAD, &t);
224	}
225
226	if (t > 0 && (size_t)t < n) {
227		n = t;
228		rn = n;
229	}
230#endif
231
232nocheck:
233	do
234		switch ((rv = read(fd, buf, n))) {
235		case -1:
236			if (errno == EINTR)
237				continue;
238			return -1;
239		case 0:
240			return rn - n;
241		default:
242			n -= rv;
243			buf = ((char *)buf) + rv;
244			break;
245		}
246	while (n > 0);
247	return rn;
248}
249
250protected int
251file_pipe2file(struct magic_set *ms, int fd, const void *startbuf,
252    size_t nbytes)
253{
254	char buf[4096];
255	ssize_t r;
256	int tfd;
257
258	(void)strlcpy(buf, "/tmp/file.XXXXXX", sizeof buf);
259#ifndef HAVE_MKSTEMP
260	{
261		char *ptr = mktemp(buf);
262		tfd = open(ptr, O_RDWR|O_TRUNC|O_EXCL|O_CREAT, 0600);
263		r = errno;
264		(void)unlink(ptr);
265		errno = r;
266	}
267#else
268	{
269		int te;
270		tfd = mkstemp(buf);
271		te = errno;
272		(void)unlink(buf);
273		errno = te;
274	}
275#endif
276	if (tfd == -1) {
277		file_error(ms, errno,
278		    "cannot create temporary file for pipe copy");
279		return -1;
280	}
281
282	if (swrite(tfd, startbuf, nbytes) != (ssize_t)nbytes)
283		r = 1;
284	else {
285		while ((r = sread(fd, buf, sizeof(buf), 1)) > 0)
286			if (swrite(tfd, buf, (size_t)r) != r)
287				break;
288	}
289
290	switch (r) {
291	case -1:
292		file_error(ms, errno, "error copying from pipe to temp file");
293		return -1;
294	case 0:
295		break;
296	default:
297		file_error(ms, errno, "error while writing to temp file");
298		return -1;
299	}
300
301	/*
302	 * We duplicate the file descriptor, because fclose on a
303	 * tmpfile will delete the file, but any open descriptors
304	 * can still access the phantom inode.
305	 */
306	if ((fd = dup2(tfd, fd)) == -1) {
307		file_error(ms, errno, "could not dup descriptor for temp file");
308		return -1;
309	}
310	(void)close(tfd);
311	if (lseek(fd, (off_t)0, SEEK_SET) == (off_t)-1) {
312		file_badseek(ms);
313		return -1;
314	}
315	return fd;
316}
317#if HAVE_FORK
318#ifdef BUILTIN_DECOMPRESS
319
320#define FHCRC		(1 << 1)
321#define FEXTRA		(1 << 2)
322#define FNAME		(1 << 3)
323#define FCOMMENT	(1 << 4)
324
325private size_t
326uncompressgzipped(struct magic_set *ms, const unsigned char *old,
327    unsigned char **newch, size_t n)
328{
329	unsigned char flg = old[3];
330	size_t data_start = 10;
331	z_stream z;
332	int rc;
333
334	if (flg & FEXTRA) {
335		if (data_start+1 >= n)
336			return 0;
337		data_start += 2 + old[data_start] + old[data_start + 1] * 256;
338	}
339	if (flg & FNAME) {
340		while(data_start < n && old[data_start])
341			data_start++;
342		data_start++;
343	}
344	if(flg & FCOMMENT) {
345		while(data_start < n && old[data_start])
346			data_start++;
347		data_start++;
348	}
349	if(flg & FHCRC)
350		data_start += 2;
351
352	if (data_start >= n)
353		return 0;
354	if ((*newch = CAST(unsigned char *, malloc(HOWMANY + 1))) == NULL) {
355		return 0;
356	}
357
358	/* XXX: const castaway, via strchr */
359	z.next_in = (Bytef *)strchr((const char *)old + data_start,
360	    old[data_start]);
361	z.avail_in = CAST(uint32_t, (n - data_start));
362	z.next_out = *newch;
363	z.avail_out = HOWMANY;
364	z.zalloc = Z_NULL;
365	z.zfree = Z_NULL;
366	z.opaque = Z_NULL;
367
368	/* LINTED bug in header macro */
369	rc = inflateInit2(&z, -15);
370	if (rc != Z_OK) {
371		file_error(ms, 0, "zlib: %s", z.msg);
372		return 0;
373	}
374
375	rc = inflate(&z, Z_SYNC_FLUSH);
376	if (rc != Z_OK && rc != Z_STREAM_END) {
377		file_error(ms, 0, "zlib: %s", z.msg);
378		return 0;
379	}
380
381	n = (size_t)z.total_out;
382	(void)inflateEnd(&z);
383
384	/* let's keep the nul-terminate tradition */
385	(*newch)[n] = '\0';
386
387	return n;
388}
389#endif
390
391private size_t
392uncompressbuf(struct magic_set *ms, int fd, size_t method,
393    const unsigned char *old, unsigned char **newch, size_t n)
394{
395	int fdin[2], fdout[2];
396	int status;
397	ssize_t r;
398
399#ifdef BUILTIN_DECOMPRESS
400        /* FIXME: This doesn't cope with bzip2 */
401	if (method == 2)
402		return uncompressgzipped(ms, old, newch, n);
403#endif
404	(void)fflush(stdout);
405	(void)fflush(stderr);
406
407	if ((fd != -1 && pipe(fdin) == -1) || pipe(fdout) == -1) {
408		file_error(ms, errno, "cannot create pipe");
409		return NODATA;
410	}
411	switch (fork()) {
412	case 0:	/* child */
413		(void) close(0);
414		if (fd != -1) {
415		    if (dup(fd) == -1)
416			_exit(1);
417		    (void) lseek(0, (off_t)0, SEEK_SET);
418		} else {
419		    if (dup(fdin[0]) == -1)
420			_exit(1);
421		    (void) close(fdin[0]);
422		    (void) close(fdin[1]);
423		}
424
425		(void) close(1);
426		if (dup(fdout[1]) == -1)
427			_exit(1);
428		(void) close(fdout[0]);
429		(void) close(fdout[1]);
430#ifndef DEBUG
431		if (compr[method].silent)
432			(void)close(2);
433#endif
434
435		(void)execvp(compr[method].argv[0],
436		    (char *const *)(intptr_t)compr[method].argv);
437#ifdef DEBUG
438		(void)fprintf(stderr, "exec `%s' failed (%s)\n",
439		    compr[method].argv[0], strerror(errno));
440#endif
441		exit(1);
442		/*NOTREACHED*/
443	case -1:
444		file_error(ms, errno, "could not fork");
445		return NODATA;
446
447	default: /* parent */
448		(void) close(fdout[1]);
449		if (fd == -1) {
450			(void) close(fdin[0]);
451			/*
452			 * fork again, to avoid blocking because both
453			 * pipes filled
454			 */
455			switch (fork()) {
456			case 0: /* child */
457				(void)close(fdout[0]);
458				if (swrite(fdin[1], old, n) != (ssize_t)n) {
459#ifdef DEBUG
460					(void)fprintf(stderr,
461					    "Write failed (%s)\n",
462					    strerror(errno));
463#endif
464					exit(1);
465				}
466				exit(0);
467				/*NOTREACHED*/
468
469			case -1:
470#ifdef DEBUG
471				(void)fprintf(stderr, "Fork failed (%s)\n",
472				    strerror(errno));
473#endif
474				exit(1);
475				/*NOTREACHED*/
476
477			default:  /* parent */
478				if (wait(&status) == -1) {
479#ifdef DEBUG
480					(void)fprintf(stderr,
481					    "Wait failed (%s)\n",
482					    strerror(errno));
483#endif
484					exit(1);
485				}
486				exit(WIFEXITED(status) ?
487				    WEXITSTATUS(status) : 1);
488				/*NOTREACHED*/
489			}
490			(void) close(fdin[1]);
491			fdin[1] = -1;
492		}
493
494		if ((*newch = (unsigned char *) malloc(HOWMANY + 1)) == NULL) {
495#ifdef DEBUG
496			(void)fprintf(stderr, "Malloc failed (%s)\n",
497			    strerror(errno));
498#endif
499			n = NODATA;
500			goto err;
501		}
502		if ((r = sread(fdout[0], *newch, HOWMANY, 0)) <= 0) {
503#ifdef DEBUG
504			(void)fprintf(stderr, "Read failed (%s)\n",
505			    strerror(errno));
506#endif
507			free(*newch);
508			n = NODATA;
509			*newch = NULL;
510			goto err;
511		} else {
512			n = r;
513		}
514 		/* NUL terminate, as every buffer is handled here. */
515 		(*newch)[n] = '\0';
516err:
517		if (fdin[1] != -1)
518			(void) close(fdin[1]);
519		(void) close(fdout[0]);
520		if (wait(&status) == -1) {
521#ifdef DEBUG
522			(void)fprintf(stderr, "Wait failed (%s)\n",
523			    strerror(errno));
524#endif
525			n = NODATA;
526		} else if (!WIFEXITED(status)) {
527#ifdef DEBUG
528			(void)fprintf(stderr, "Child not exited (0x%x)\n",
529			    status);
530#endif
531		} else if (WEXITSTATUS(status) != 0) {
532#ifdef DEBUG
533			(void)fprintf(stderr, "Child exited (0x%d)\n",
534			    WEXITSTATUS(status));
535#endif
536		}
537
538		(void) close(fdin[0]);
539
540		return n;
541	}
542}
543#endif
544