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 1.100 2016/10/24 18:02:17 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)
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.
87 static const char zlibcode[] =
88 "import sys, zlib; sys.stdout.write(zlib.decompress(sys.stdin.read()))";
90 static const char *zlib_args[] = { "python", "-c", zlibcode, NULL };
93 zlibcmp(const unsigned char *buf)
96 unsigned char *s = (unsigned char *)&x;
98 if ((buf[0] & 0xf) != 8 || (buf[0] & 0x80) != 0)
100 if (s[0] != 1) /* endianness test */
101 x = buf[0] | (buf[1] << 8);
103 x = buf[1] | (buf[0] << 8);
110 #define gzip_flags "-cd"
111 #define lrzip_flags "-do"
112 #define lzip_flags gzip_flags
114 static const char *gzip_args[] = {
115 "gzip", gzip_flags, NULL
117 static const char *uncompress_args[] = {
118 "uncompress", "-c", NULL
120 static const char *bzip2_args[] = {
123 static const char *lzip_args[] = {
124 "lzip", lzip_flags, NULL
126 static const char *xz_args[] = {
129 static const char *lrzip_args[] = {
130 "lrzip", lrzip_flags, NULL
132 static const char *lz4_args[] = {
135 static const char *zstd_args[] = {
139 private const struct {
144 { "\037\235", 2, gzip_args }, /* compressed */
145 /* Uncompress can get stuck; so use gzip first if we have it
146 * Idea from Damien Clark, thanks! */
147 { "\037\235", 2, uncompress_args }, /* compressed */
148 { "\037\213", 2, gzip_args }, /* gzipped */
149 { "\037\236", 2, gzip_args }, /* frozen */
150 { "\037\240", 2, gzip_args }, /* SCO LZH */
151 /* the standard pack utilities do not accept standard input */
152 { "\037\036", 2, gzip_args }, /* packed */
153 { "PK\3\4", 4, gzip_args }, /* pkzipped, */
154 /* ...only first file examined */
155 { "BZh", 3, bzip2_args }, /* bzip2-ed */
156 { "LZIP", 4, lzip_args }, /* lzip-ed */
157 { "\3757zXZ\0", 6, xz_args }, /* XZ Utils */
158 { "LRZI", 4, lrzip_args }, /* LRZIP */
159 { "\004\"M\030",4, lz4_args }, /* LZ4 */
160 { "\x28\xB5\x2F\xFD", 4, zstd_args }, /* zstd */
162 { RCAST(const void *, zlibcmp), 0, zlib_args }, /* zlib */
170 private ssize_t swrite(int, const void *, size_t);
172 private size_t ncompr = sizeof(compr) / sizeof(compr[0]);
173 private int uncompressbuf(int, size_t, size_t, const unsigned char *,
174 unsigned char **, size_t *);
175 #ifdef BUILTIN_DECOMPRESS
176 private int uncompresszlib(const unsigned char *, unsigned char **, size_t,
178 private int uncompressgzipped(const unsigned char *, unsigned char **, size_t,
181 static int makeerror(unsigned char **, size_t *, const char *, ...)
182 __attribute__((__format__(__printf__, 3, 4)));
183 private const char *methodname(size_t);
186 file_zmagic(struct magic_set *ms, int fd, const char *name,
187 const unsigned char *buf, size_t nbytes)
189 unsigned char *newbuf = NULL;
193 int urv, prv, rv = 0;
194 int mime = ms->flags & MAGIC_MIME;
199 if ((ms->flags & MAGIC_COMPRESS) == 0)
203 osigpipe = signal(SIGPIPE, SIG_IGN);
205 for (i = 0; i < ncompr; i++) {
207 if (nbytes < compr[i].maglen)
210 if (compr[i].maglen == 0)
211 zm = (RCAST(int (*)(const unsigned char *),
212 CCAST(void *, compr[i].magic)))(buf);
215 zm = memcmp(buf, compr[i].magic, compr[i].maglen) == 0;
220 urv = uncompressbuf(fd, ms->bytes_max, i, buf, &newbuf, &nsz);
221 DPRINTF("uncompressbuf = %d, %s, %zu\n", urv, (char *)newbuf,
227 ms->flags &= ~MAGIC_COMPRESS;
229 prv = file_printf(ms, "%s ERROR: %s",
230 methodname(i), newbuf);
232 prv = file_buffer(ms, -1, name, newbuf, nsz);
236 if ((ms->flags & MAGIC_COMPRESS_TRANSP) != 0)
238 if (mime != MAGIC_MIME && mime != 0)
241 mime ? " compressed-encoding=" : " (")) == -1)
243 if ((pb = file_push_buffer(ms)) == NULL)
246 * XXX: If file_buffer fails here, we overwrite
247 * the compressed text. FIXME.
249 if (file_buffer(ms, -1, NULL, buf, nbytes) == -1)
251 if ((rbuf = file_pop_buffer(ms, pb)) != NULL) {
252 if (file_printf(ms, "%s", rbuf) == -1) {
258 if (!mime && file_printf(ms, ")") == -1)
272 DPRINTF("rv = %d\n", rv);
275 (void)signal(SIGPIPE, osigpipe);
278 ms->flags |= MAGIC_COMPRESS;
279 DPRINTF("Zmagic returns %d\n", rv);
284 * `safe' write for sockets and pipes.
287 swrite(int fd, const void *buf, size_t n)
293 switch (rv = write(fd, buf, n)) {
300 buf = CAST(const char *, buf) + rv;
309 * `safe' read for sockets and pipes.
312 sread(int fd, void *buf, size_t n, int canbepipe __attribute__((__unused__)))
320 if (fd == STDIN_FILENO)
324 if (canbepipe && (ioctl(fd, FIONREAD, &t) == -1 || t == 0)) {
327 for (cnt = 0;; cnt++) {
329 struct timeval tout = {0, 100 * 1000};
336 * Avoid soft deadlock: do not read if there
337 * is nothing to read from sockets and pipes.
339 selrv = select(fd + 1, &check, NULL, NULL, &tout);
341 if (errno == EINTR || errno == EAGAIN)
343 } else if (selrv == 0 && cnt >= 5) {
349 (void)ioctl(fd, FIONREAD, &t);
352 if (t > 0 && (size_t)t < n) {
360 switch ((rv = read(fd, buf, n))) {
369 buf = CAST(char *, CCAST(void *, buf)) + rv;
377 file_pipe2file(struct magic_set *ms, int fd, const void *startbuf,
384 (void)strlcpy(buf, "/tmp/file.XXXXXX", sizeof buf);
387 char *ptr = mktemp(buf);
388 tfd = open(ptr, O_RDWR|O_TRUNC|O_EXCL|O_CREAT, 0600);
403 file_error(ms, errno,
404 "cannot create temporary file for pipe copy");
408 if (swrite(tfd, startbuf, nbytes) != (ssize_t)nbytes)
411 while ((r = sread(fd, buf, sizeof(buf), 1)) > 0)
412 if (swrite(tfd, buf, (size_t)r) != r)
418 file_error(ms, errno, "error copying from pipe to temp file");
423 file_error(ms, errno, "error while writing to temp file");
428 * We duplicate the file descriptor, because fclose on a
429 * tmpfile will delete the file, but any open descriptors
430 * can still access the phantom inode.
432 if ((fd = dup2(tfd, fd)) == -1) {
433 file_error(ms, errno, "could not dup descriptor for temp file");
437 if (lseek(fd, (off_t)0, SEEK_SET) == (off_t)-1) {
444 #ifdef BUILTIN_DECOMPRESS
446 #define FHCRC (1 << 1)
447 #define FEXTRA (1 << 2)
448 #define FNAME (1 << 3)
449 #define FCOMMENT (1 << 4)
453 uncompressgzipped(const unsigned char *old, unsigned char **newch,
454 size_t bytes_max, size_t *n)
456 unsigned char flg = old[3];
457 size_t data_start = 10;
460 if (data_start + 1 >= *n)
462 data_start += 2 + old[data_start] + old[data_start + 1] * 256;
465 while(data_start < *n && old[data_start])
469 if (flg & FCOMMENT) {
470 while(data_start < *n && old[data_start])
477 if (data_start >= *n)
482 return uncompresszlib(old, newch, bytes_max, n, 0);
484 return makeerror(newch, n, "File too short");
488 uncompresszlib(const unsigned char *old, unsigned char **newch,
489 size_t bytes_max, size_t *n, int zlib)
494 if ((*newch = CAST(unsigned char *, malloc(bytes_max + 1))) == NULL)
495 return makeerror(newch, n, "No buffer, %s", strerror(errno));
497 z.next_in = CCAST(Bytef *, old);
498 z.avail_in = CAST(uint32_t, *n);
500 z.avail_out = bytes_max;
505 /* LINTED bug in header macro */
506 rc = zlib ? inflateInit(&z) : inflateInit2(&z, -15);
510 rc = inflate(&z, Z_SYNC_FLUSH);
511 if (rc != Z_OK && rc != Z_STREAM_END)
514 *n = (size_t)z.total_out;
519 /* let's keep the nul-terminate tradition */
524 strlcpy((char *)*newch, z.msg ? z.msg : zError(rc), bytes_max);
525 *n = strlen((char *)*newch);
531 makeerror(unsigned char **buf, size_t *len, const char *fmt, ...)
538 rv = vasprintf(&msg, fmt, ap);
545 *buf = (unsigned char *)msg;
551 closefd(int *fd, size_t i)
563 for (i = 0; i < 2; i++)
568 copydesc(int i, int *fd)
570 int j = fd[i == STDIN_FILENO ? 0 : 1];
573 if (dup2(j, i) == -1) {
574 DPRINTF("dup(%d, %d) failed (%s)\n", j, i, strerror(errno));
581 writechild(int fdp[3][2], const void *old, size_t n)
585 closefd(fdp[STDIN_FILENO], 0);
587 * fork again, to avoid blocking because both
592 closefd(fdp[STDOUT_FILENO], 0);
593 if (swrite(fdp[STDIN_FILENO][1], old, n) != (ssize_t)n) {
594 DPRINTF("Write failed (%s)\n", strerror(errno));
601 DPRINTF("Fork failed (%s)\n", strerror(errno));
605 default: /* parent */
606 if (wait(&status) == -1) {
607 DPRINTF("Wait failed (%s)\n", strerror(errno));
610 DPRINTF("Grandchild wait return %#x\n", status);
612 closefd(fdp[STDIN_FILENO], 1);
616 filter_error(unsigned char *ubuf, ssize_t n)
623 while (isspace((unsigned char)*buf))
625 DPRINTF("Filter error[[[%s]]]\n", buf);
626 if ((p = strchr((char *)buf, '\n')) != NULL)
628 if ((p = strchr((char *)buf, ';')) != NULL)
630 if ((p = strrchr((char *)buf, ':')) != NULL) {
632 while (isspace((unsigned char)*p))
635 memmove(ubuf, p, n + 1);
637 DPRINTF("Filter error after[[[%s]]]\n", (char *)ubuf);
639 *ubuf = toupper(*ubuf);
644 methodname(size_t method)
646 #ifdef BUILTIN_DECOMPRESS
647 /* FIXME: This doesn't cope with bzip2 */
648 if (method == 2 || compr[method].maglen == 0)
651 return compr[method].argv[0];
655 uncompressbuf(int fd, size_t bytes_max, size_t method, const unsigned char *old,
656 unsigned char **newch, size_t* n)
663 #ifdef BUILTIN_DECOMPRESS
664 /* FIXME: This doesn't cope with bzip2 */
666 return uncompressgzipped(old, newch, bytes_max, n);
667 if (compr[method].maglen == 0)
668 return uncompresszlib(old, newch, bytes_max, n, 1);
670 (void)fflush(stdout);
671 (void)fflush(stderr);
673 for (i = 0; i < __arraycount(fdp); i++)
674 fdp[i][0] = fdp[i][1] = -1;
676 if ((fd == -1 && pipe(fdp[STDIN_FILENO]) == -1) ||
677 pipe(fdp[STDOUT_FILENO]) == -1 || pipe(fdp[STDERR_FILENO]) == -1) {
678 closep(fdp[STDIN_FILENO]);
679 closep(fdp[STDOUT_FILENO]);
680 return makeerror(newch, n, "Cannot create pipe, %s",
686 fdp[STDIN_FILENO][0] = fd;
687 (void) lseek(fd, (off_t)0, SEEK_SET);
690 for (i = 0; i < __arraycount(fdp); i++)
693 (void)execvp(compr[method].argv[0],
694 (char *const *)(intptr_t)compr[method].argv);
695 dprintf(STDERR_FILENO, "exec `%s' failed, %s",
696 compr[method].argv[0], strerror(errno));
700 return makeerror(newch, n, "Cannot fork, %s",
703 default: /* parent */
704 for (i = 1; i < __arraycount(fdp); i++)
707 /* Write the buffer data to the child, if we don't have fd */
709 writechild(fdp, old, *n);
711 *newch = CAST(unsigned char *, malloc(bytes_max + 1));
712 if (*newch == NULL) {
713 rv = makeerror(newch, n, "No buffer, %s",
718 if ((r = sread(fdp[STDOUT_FILENO][0], *newch, bytes_max, 0)) > 0)
720 DPRINTF("Read stdout failed %d (%s)\n", fdp[STDOUT_FILENO][0],
721 r != -1 ? strerror(errno) : "no data");
725 (r = sread(fdp[STDERR_FILENO][0], *newch, bytes_max, 0)) > 0)
727 r = filter_error(*newch, r);
732 rv = makeerror(newch, n, "Read failed, %s",
735 rv = makeerror(newch, n, "No data");
740 /* NUL terminate, as every buffer is handled here. */
743 closefd(fdp[STDIN_FILENO], 1);
744 closefd(fdp[STDOUT_FILENO], 0);
745 closefd(fdp[STDERR_FILENO], 0);
746 if (wait(&status) == -1) {
748 rv = makeerror(newch, n, "Wait failed, %s", strerror(errno));
749 DPRINTF("Child wait return %#x\n", status);
750 } else if (!WIFEXITED(status)) {
751 DPRINTF("Child not exited (0x%x)\n", status);
752 } else if (WEXITSTATUS(status) != 0) {
753 DPRINTF("Child exited (0x%d)\n", WEXITSTATUS(status));
756 closefd(fdp[STDIN_FILENO], 0);
757 DPRINTF("Returning %p n=%zu rv=%d\n", *newch, *n, rv);