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.
6 * Redistribution and use in source and binary forms, with or without
7 * modification, are permitted provided that the following conditions
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.
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
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
38 FILE_RCSID("@(#)$File: compress.c,v 1.106 2017/11/02 20:25:39 christos Exp $")
53 typedef void (*sig_t)(int);
54 # endif /* HAVE_SIG_T */
56 #if !defined(__MINGW32__) && !defined(WIN32)
57 #include <sys/ioctl.h>
59 #ifdef HAVE_SYS_WAIT_H
62 #if defined(HAVE_SYS_TIME_H)
65 #if defined(HAVE_ZLIB_H) && defined(ZLIBSUPPORT)
66 #define BUILTIN_DECOMPRESS
71 #define DPRINTF(...) do { \
73 tty = open("/dev/tty", O_RDWR); \
76 dprintf(tty, __VA_ARGS__); \
77 } while (/*CONSTCOND*/0)
84 * The following python code is not really used because ZLIBSUPPORT is only
85 * defined if we have a built-in zlib, and the built-in zlib handles that.
86 * That is not true for android where we have zlib.h and not -lz.
88 static const char zlibcode[] =
89 "import sys, zlib; sys.stdout.write(zlib.decompress(sys.stdin.read()))";
91 static const char *zlib_args[] = { "python", "-c", zlibcode, NULL };
94 zlibcmp(const unsigned char *buf)
97 unsigned char *s = CAST(unsigned char *, CAST(void *, &x));
99 if ((buf[0] & 0xf) != 8 || (buf[0] & 0x80) != 0)
101 if (s[0] != 1) /* endianness test */
102 x = buf[0] | (buf[1] << 8);
104 x = buf[1] | (buf[0] << 8);
111 #define gzip_flags "-cd"
112 #define lrzip_flags "-do"
113 #define lzip_flags gzip_flags
115 static const char *gzip_args[] = {
116 "gzip", gzip_flags, NULL
118 static const char *uncompress_args[] = {
119 "uncompress", "-c", NULL
121 static const char *bzip2_args[] = {
124 static const char *lzip_args[] = {
125 "lzip", lzip_flags, NULL
127 static const char *xz_args[] = {
130 static const char *lrzip_args[] = {
131 "lrzip", lrzip_flags, NULL
133 static const char *lz4_args[] = {
136 static const char *zstd_args[] = {
140 private const struct {
145 { "\037\235", 2, gzip_args }, /* compressed */
146 /* Uncompress can get stuck; so use gzip first if we have it
147 * Idea from Damien Clark, thanks! */
148 { "\037\235", 2, uncompress_args }, /* compressed */
149 { "\037\213", 2, gzip_args }, /* gzipped */
150 { "\037\236", 2, gzip_args }, /* frozen */
151 { "\037\240", 2, gzip_args }, /* SCO LZH */
152 /* the standard pack utilities do not accept standard input */
153 { "\037\036", 2, gzip_args }, /* packed */
154 { "PK\3\4", 4, gzip_args }, /* pkzipped, */
155 /* ...only first file examined */
156 { "BZh", 3, bzip2_args }, /* bzip2-ed */
157 { "LZIP", 4, lzip_args }, /* lzip-ed */
158 { "\3757zXZ\0", 6, xz_args }, /* XZ Utils */
159 { "LRZI", 4, lrzip_args }, /* LRZIP */
160 { "\004\"M\030",4, lz4_args }, /* LZ4 */
161 { "\x28\xB5\x2F\xFD", 4, zstd_args }, /* zstd */
163 { RCAST(const void *, zlibcmp), 0, zlib_args }, /* zlib */
171 private ssize_t swrite(int, const void *, size_t);
173 private size_t ncompr = sizeof(compr) / sizeof(compr[0]);
174 private int uncompressbuf(int, size_t, size_t, const unsigned char *,
175 unsigned char **, size_t *);
176 #ifdef BUILTIN_DECOMPRESS
177 private int uncompresszlib(const unsigned char *, unsigned char **, size_t,
179 private int uncompressgzipped(const unsigned char *, unsigned char **, size_t,
182 static int makeerror(unsigned char **, size_t *, const char *, ...)
183 __attribute__((__format__(__printf__, 3, 4)));
184 private const char *methodname(size_t);
187 file_zmagic(struct magic_set *ms, const struct buffer *b, const char *name)
189 unsigned char *newbuf = NULL;
193 int urv, prv, rv = 0;
194 int mime = ms->flags & MAGIC_MIME;
196 const unsigned char *buf = b->fbuf;
197 size_t nbytes = b->flen;
202 if ((ms->flags & MAGIC_COMPRESS) == 0)
206 osigpipe = signal(SIGPIPE, SIG_IGN);
208 for (i = 0; i < ncompr; i++) {
210 if (nbytes < compr[i].maglen)
213 if (compr[i].maglen == 0)
214 zm = (RCAST(int (*)(const unsigned char *),
215 CCAST(void *, compr[i].magic)))(buf);
218 zm = memcmp(buf, compr[i].magic, compr[i].maglen) == 0;
223 urv = uncompressbuf(fd, ms->bytes_max, i, buf, &newbuf, &nsz);
224 DPRINTF("uncompressbuf = %d, %s, %zu\n", urv, (char *)newbuf,
230 ms->flags &= ~MAGIC_COMPRESS;
232 prv = file_printf(ms, "%s ERROR: %s",
233 methodname(i), newbuf);
235 prv = file_buffer(ms, -1, name, newbuf, nsz);
239 if ((ms->flags & MAGIC_COMPRESS_TRANSP) != 0)
241 if (mime != MAGIC_MIME && mime != 0)
244 mime ? " compressed-encoding=" : " (")) == -1)
246 if ((pb = file_push_buffer(ms)) == NULL)
249 * XXX: If file_buffer fails here, we overwrite
250 * the compressed text. FIXME.
252 if (file_buffer(ms, -1, NULL, buf, nbytes) == -1)
254 if ((rbuf = file_pop_buffer(ms, pb)) != NULL) {
255 if (file_printf(ms, "%s", rbuf) == -1) {
261 if (!mime && file_printf(ms, ")") == -1)
275 DPRINTF("rv = %d\n", rv);
278 (void)signal(SIGPIPE, osigpipe);
281 ms->flags |= MAGIC_COMPRESS;
282 DPRINTF("Zmagic returns %d\n", rv);
287 * `safe' write for sockets and pipes.
290 swrite(int fd, const void *buf, size_t n)
296 switch (rv = write(fd, buf, n)) {
303 buf = CAST(const char *, buf) + rv;
312 * `safe' read for sockets and pipes.
315 sread(int fd, void *buf, size_t n, int canbepipe __attribute__((__unused__)))
323 if (fd == STDIN_FILENO)
327 if (canbepipe && (ioctl(fd, FIONREAD, &t) == -1 || t == 0)) {
330 for (cnt = 0;; cnt++) {
332 struct timeval tout = {0, 100 * 1000};
339 * Avoid soft deadlock: do not read if there
340 * is nothing to read from sockets and pipes.
342 selrv = select(fd + 1, &check, NULL, NULL, &tout);
344 if (errno == EINTR || errno == EAGAIN)
346 } else if (selrv == 0 && cnt >= 5) {
352 (void)ioctl(fd, FIONREAD, &t);
355 if (t > 0 && (size_t)t < n) {
363 switch ((rv = read(fd, buf, n))) {
372 buf = CAST(char *, CCAST(void *, buf)) + rv;
380 file_pipe2file(struct magic_set *ms, int fd, const void *startbuf,
387 (void)strlcpy(buf, "/tmp/file.XXXXXX", sizeof buf);
390 char *ptr = mktemp(buf);
391 tfd = open(ptr, O_RDWR|O_TRUNC|O_EXCL|O_CREAT, 0600);
406 file_error(ms, errno,
407 "cannot create temporary file for pipe copy");
411 if (swrite(tfd, startbuf, nbytes) != (ssize_t)nbytes)
414 while ((r = sread(fd, buf, sizeof(buf), 1)) > 0)
415 if (swrite(tfd, buf, (size_t)r) != r)
421 file_error(ms, errno, "error copying from pipe to temp file");
426 file_error(ms, errno, "error while writing to temp file");
431 * We duplicate the file descriptor, because fclose on a
432 * tmpfile will delete the file, but any open descriptors
433 * can still access the phantom inode.
435 if ((fd = dup2(tfd, fd)) == -1) {
436 file_error(ms, errno, "could not dup descriptor for temp file");
440 if (lseek(fd, (off_t)0, SEEK_SET) == (off_t)-1) {
447 #ifdef BUILTIN_DECOMPRESS
449 #define FHCRC (1 << 1)
450 #define FEXTRA (1 << 2)
451 #define FNAME (1 << 3)
452 #define FCOMMENT (1 << 4)
456 uncompressgzipped(const unsigned char *old, unsigned char **newch,
457 size_t bytes_max, size_t *n)
459 unsigned char flg = old[3];
460 size_t data_start = 10;
463 if (data_start + 1 >= *n)
465 data_start += 2 + old[data_start] + old[data_start + 1] * 256;
468 while(data_start < *n && old[data_start])
472 if (flg & FCOMMENT) {
473 while(data_start < *n && old[data_start])
480 if (data_start >= *n)
485 return uncompresszlib(old, newch, bytes_max, n, 0);
487 return makeerror(newch, n, "File too short");
491 uncompresszlib(const unsigned char *old, unsigned char **newch,
492 size_t bytes_max, size_t *n, int zlib)
497 if ((*newch = CAST(unsigned char *, malloc(bytes_max + 1))) == NULL)
498 return makeerror(newch, n, "No buffer, %s", strerror(errno));
500 z.next_in = CCAST(Bytef *, old);
501 z.avail_in = CAST(uint32_t, *n);
503 z.avail_out = CAST(unsigned int, bytes_max);
508 /* LINTED bug in header macro */
509 rc = zlib ? inflateInit(&z) : inflateInit2(&z, -15);
513 rc = inflate(&z, Z_SYNC_FLUSH);
514 if (rc != Z_OK && rc != Z_STREAM_END)
517 *n = (size_t)z.total_out;
522 /* let's keep the nul-terminate tradition */
527 strlcpy((char *)*newch, z.msg ? z.msg : zError(rc), bytes_max);
528 *n = strlen((char *)*newch);
534 makeerror(unsigned char **buf, size_t *len, const char *fmt, ...)
541 rv = vasprintf(&msg, fmt, ap);
548 *buf = (unsigned char *)msg;
554 closefd(int *fd, size_t i)
566 for (i = 0; i < 2; i++)
571 copydesc(int i, int *fd)
573 int j = fd[i == STDIN_FILENO ? 0 : 1];
576 if (dup2(j, i) == -1) {
577 DPRINTF("dup(%d, %d) failed (%s)\n", j, i, strerror(errno));
584 writechild(int fdp[3][2], const void *old, size_t n)
588 closefd(fdp[STDIN_FILENO], 0);
590 * fork again, to avoid blocking because both
595 closefd(fdp[STDOUT_FILENO], 0);
596 if (swrite(fdp[STDIN_FILENO][1], old, n) != (ssize_t)n) {
597 DPRINTF("Write failed (%s)\n", strerror(errno));
604 DPRINTF("Fork failed (%s)\n", strerror(errno));
608 default: /* parent */
609 if (wait(&status) == -1) {
610 DPRINTF("Wait failed (%s)\n", strerror(errno));
613 DPRINTF("Grandchild wait return %#x\n", status);
615 closefd(fdp[STDIN_FILENO], 1);
619 filter_error(unsigned char *ubuf, ssize_t n)
626 while (isspace((unsigned char)*buf))
628 DPRINTF("Filter error[[[%s]]]\n", buf);
629 if ((p = strchr((char *)buf, '\n')) != NULL)
631 if ((p = strchr((char *)buf, ';')) != NULL)
633 if ((p = strrchr((char *)buf, ':')) != NULL) {
635 while (isspace((unsigned char)*p))
638 memmove(ubuf, p, CAST(size_t, n + 1));
640 DPRINTF("Filter error after[[[%s]]]\n", (char *)ubuf);
642 *ubuf = toupper(*ubuf);
647 methodname(size_t method)
649 #ifdef BUILTIN_DECOMPRESS
650 /* FIXME: This doesn't cope with bzip2 */
651 if (method == 2 || compr[method].maglen == 0)
654 return compr[method].argv[0];
658 uncompressbuf(int fd, size_t bytes_max, size_t method, const unsigned char *old,
659 unsigned char **newch, size_t* n)
666 #ifdef BUILTIN_DECOMPRESS
667 /* FIXME: This doesn't cope with bzip2 */
669 return uncompressgzipped(old, newch, bytes_max, n);
670 if (compr[method].maglen == 0)
671 return uncompresszlib(old, newch, bytes_max, n, 1);
673 (void)fflush(stdout);
674 (void)fflush(stderr);
676 for (i = 0; i < __arraycount(fdp); i++)
677 fdp[i][0] = fdp[i][1] = -1;
679 if ((fd == -1 && pipe(fdp[STDIN_FILENO]) == -1) ||
680 pipe(fdp[STDOUT_FILENO]) == -1 || pipe(fdp[STDERR_FILENO]) == -1) {
681 closep(fdp[STDIN_FILENO]);
682 closep(fdp[STDOUT_FILENO]);
683 return makeerror(newch, n, "Cannot create pipe, %s",
689 fdp[STDIN_FILENO][0] = fd;
690 (void) lseek(fd, (off_t)0, SEEK_SET);
693 for (i = 0; i < __arraycount(fdp); i++)
694 copydesc(CAST(int, i), fdp[i]);
696 (void)execvp(compr[method].argv[0],
697 (char *const *)(intptr_t)compr[method].argv);
698 dprintf(STDERR_FILENO, "exec `%s' failed, %s",
699 compr[method].argv[0], strerror(errno));
703 return makeerror(newch, n, "Cannot fork, %s",
706 default: /* parent */
707 for (i = 1; i < __arraycount(fdp); i++)
710 /* Write the buffer data to the child, if we don't have fd */
712 writechild(fdp, old, *n);
714 *newch = CAST(unsigned char *, malloc(bytes_max + 1));
715 if (*newch == NULL) {
716 rv = makeerror(newch, n, "No buffer, %s",
721 if ((r = sread(fdp[STDOUT_FILENO][0], *newch, bytes_max, 0)) > 0)
723 DPRINTF("Read stdout failed %d (%s)\n", fdp[STDOUT_FILENO][0],
724 r != -1 ? strerror(errno) : "no data");
728 (r = sread(fdp[STDERR_FILENO][0], *newch, bytes_max, 0)) > 0)
730 r = filter_error(*newch, r);
735 rv = makeerror(newch, n, "Read failed, %s",
738 rv = makeerror(newch, n, "No data");
743 /* NUL terminate, as every buffer is handled here. */
746 closefd(fdp[STDIN_FILENO], 1);
747 closefd(fdp[STDOUT_FILENO], 0);
748 closefd(fdp[STDERR_FILENO], 0);
749 if (wait(&status) == -1) {
751 rv = makeerror(newch, n, "Wait failed, %s", strerror(errno));
752 DPRINTF("Child wait return %#x\n", status);
753 } else if (!WIFEXITED(status)) {
754 DPRINTF("Child not exited (%#x)\n", status);
755 } else if (WEXITSTATUS(status) != 0) {
756 DPRINTF("Child exited (%#x)\n", WEXITSTATUS(status));
759 closefd(fdp[STDIN_FILENO], 0);
760 DPRINTF("Returning %p n=%zu rv=%d\n", *newch, *n, rv);