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 "config.h"
36 #include "file.h"
37
38 #ifndef lint
39 FILE_RCSID("@(#)$File: compress.c,v 1.77 2014/12/12 16:33:01 christos Exp $")
40 #endif
41
42 #include "magic.h"
43 #include <stdlib.h>
44 #ifdef HAVE_UNISTD_H
45 #include <unistd.h>
46 #endif
47 #include <string.h>
48 #include <errno.h>
49 #ifdef HAVE_SIGNAL_H
50 #include <signal.h>
51 # ifndef HAVE_SIG_T
52 typedef void (*sig_t)(int);
53 # endif /* HAVE_SIG_T */
54 #endif
55 #ifndef PHP_WIN32
56 #include <sys/ioctl.h>
57 #endif
58 #ifdef HAVE_SYS_WAIT_H
59 #include <sys/wait.h>
60 #endif
61 #if defined(HAVE_SYS_TIME_H)
62 #include <sys/time.h>
63 #endif
64 #if defined(HAVE_ZLIB_H) && defined(HAVE_LIBZ)
65 #define BUILTIN_DECOMPRESS
66 #include <zlib.h>
67 #endif
68
69 #undef FIONREAD
70
71
72 private const struct {
73 const char magic[8];
74 size_t maglen;
75 const char *argv[3];
76 int silent;
77 } compr[] = {
78 { "\037\235", 2, { "gzip", "-cdq", NULL }, 1 }, /* compressed */
79 /* Uncompress can get stuck; so use gzip first if we have it
80 * Idea from Damien Clark, thanks! */
81 { "\037\235", 2, { "uncompress", "-c", NULL }, 1 }, /* compressed */
82 { "\037\213", 2, { "gzip", "-cdq", NULL }, 1 }, /* gzipped */
83 { "\037\236", 2, { "gzip", "-cdq", NULL }, 1 }, /* frozen */
84 { "\037\240", 2, { "gzip", "-cdq", NULL }, 1 }, /* SCO LZH */
85 /* the standard pack utilities do not accept standard input */
86 { "\037\036", 2, { "gzip", "-cdq", NULL }, 0 }, /* packed */
87 { "PK\3\4", 4, { "gzip", "-cdq", NULL }, 1 }, /* pkzipped, */
88 /* ...only first file examined */
89 { "BZh", 3, { "bzip2", "-cd", NULL }, 1 }, /* bzip2-ed */
90 { "LZIP", 4, { "lzip", "-cdq", NULL }, 1 },
91 { "\3757zXZ\0",6,{ "xz", "-cd", NULL }, 1 }, /* XZ Utils */
92 { "LRZI", 4, { "lrzip", "-dqo-", NULL }, 1 }, /* LRZIP */
93 { "\004\"M\030", 4, { "lz4", "-cd", NULL }, 1 }, /* LZ4 */
94 };
95
96 #define NODATA ((size_t)~0)
97
98 private ssize_t swrite(int, const void *, size_t);
99 #ifdef PHP_FILEINFO_UNCOMPRESS
100 private size_t uncompressbuf(struct magic_set *, int, size_t,
101 const unsigned char *, unsigned char **, size_t);
102 #ifdef BUILTIN_DECOMPRESS
103 private size_t uncompressgzipped(struct magic_set *, const unsigned char *,
104 unsigned char **, size_t);
105 #endif
106
107 protected int
file_zmagic(struct magic_set * ms,int fd,const char * name,const unsigned char * buf,size_t nbytes)108 file_zmagic(struct magic_set *ms, int fd, const char *name,
109 const unsigned char *buf, size_t nbytes)
110 {
111 unsigned char *newbuf = NULL;
112 size_t i, nsz;
113 int rv = 0;
114 int mime = ms->flags & MAGIC_MIME;
115 #ifdef HAVE_SIGNAL_H
116 sig_t osigpipe;
117 #endif
118
119 if ((ms->flags & MAGIC_COMPRESS) == 0)
120 return 0;
121
122 #ifdef HAVE_SIGNAL_H
123 osigpipe = signal(SIGPIPE, SIG_IGN);
124 #endif
125 for (i = 0; i < ncompr; i++) {
126 if (nbytes < compr[i].maglen)
127 continue;
128 if (memcmp(buf, compr[i].magic, compr[i].maglen) == 0 &&
129 (nsz = uncompressbuf(ms, fd, i, buf, &newbuf,
130 nbytes)) != NODATA) {
131 ms->flags &= ~MAGIC_COMPRESS;
132 rv = -1;
133 if (file_buffer(ms, -1, name, newbuf, nsz) == -1)
134 goto error;
135
136 if (mime == MAGIC_MIME || mime == 0) {
137 if (file_printf(ms, mime ?
138 " compressed-encoding=" : " (") == -1)
139 goto error;
140 if (file_buffer(ms, -1, NULL, buf, nbytes) == -1)
141 goto error;
142 if (!mime && file_printf(ms, ")") == -1)
143 goto error;
144 }
145
146 rv = 1;
147 break;
148 }
149 }
150 error:
151 #ifdef HAVE_SIGNAL_H
152 (void)signal(SIGPIPE, osigpipe);
153 #endif
154 if (newbuf)
155 efree(newbuf);
156 ms->flags |= MAGIC_COMPRESS;
157 return rv;
158 }
159 #endif
160 /*
161 * `safe' write for sockets and pipes.
162 */
163 private ssize_t
swrite(int fd,const void * buf,size_t n)164 swrite(int fd, const void *buf, size_t n)
165 {
166 ssize_t rv;
167 size_t rn = n;
168
169 do
170 switch (rv = write(fd, buf, n)) {
171 case -1:
172 if (errno == EINTR)
173 continue;
174 return -1;
175 default:
176 n -= rv;
177 buf = CAST(const char *, buf) + rv;
178 break;
179 }
180 while (n > 0);
181 return rn;
182 }
183
184
185 /*
186 * `safe' read for sockets and pipes.
187 */
188 protected ssize_t
sread(int fd,void * buf,size_t n,int canbepipe)189 sread(int fd, void *buf, size_t n, int canbepipe)
190 {
191 ssize_t rv;
192 #ifdef FIONREAD
193 int t = 0;
194 #endif
195 size_t rn = n;
196
197 if (fd == STDIN_FILENO)
198 goto nocheck;
199
200 #ifdef FIONREAD
201 if (canbepipe && (ioctl(fd, FIONREAD, &t) == -1 || t == 0)) {
202 #ifdef FD_ZERO
203 ssize_t cnt;
204 for (cnt = 0;; cnt++) {
205 fd_set check;
206 struct timeval tout = {0, 100 * 1000};
207 int selrv;
208
209 FD_ZERO(&check);
210 FD_SET(fd, &check);
211
212 /*
213 * Avoid soft deadlock: do not read if there
214 * is nothing to read from sockets and pipes.
215 */
216 selrv = select(fd + 1, &check, NULL, NULL, &tout);
217 if (selrv == -1) {
218 if (errno == EINTR || errno == EAGAIN)
219 continue;
220 } else if (selrv == 0 && cnt >= 5) {
221 return 0;
222 } else
223 break;
224 }
225 #endif
226 (void)ioctl(fd, FIONREAD, &t);
227 }
228
229 if (t > 0 && (size_t)t < n) {
230 n = t;
231 rn = n;
232 }
233 #endif
234
235 nocheck:
236 do
237 switch ((rv = FINFO_READ_FUNC(fd, buf, n))) {
238 case -1:
239 if (errno == EINTR)
240 continue;
241 return -1;
242 case 0:
243 return rn - n;
244 default:
245 n -= rv;
246 buf = ((char *)buf) + rv;
247 break;
248 }
249 while (n > 0);
250 return rn;
251 }
252
253 protected int
file_pipe2file(struct magic_set * ms,int fd,const void * startbuf,size_t nbytes)254 file_pipe2file(struct magic_set *ms, int fd, const void *startbuf,
255 size_t nbytes)
256 {
257 char buf[4096];
258 ssize_t r;
259 int tfd;
260
261 (void)strlcpy(buf, "/tmp/file.XXXXXX", sizeof buf);
262 #ifndef HAVE_MKSTEMP
263 {
264 char *ptr = mktemp(buf);
265 tfd = open(ptr, O_RDWR|O_TRUNC|O_EXCL|O_CREAT, 0600);
266 r = errno;
267 (void)unlink(ptr);
268 errno = r;
269 }
270 #else
271 {
272 int te;
273 tfd = mkstemp(buf);
274 te = errno;
275 (void)unlink(buf);
276 errno = te;
277 }
278 #endif
279 if (tfd == -1) {
280 file_error(ms, errno,
281 "cannot create temporary file for pipe copy");
282 return -1;
283 }
284
285 if (swrite(tfd, startbuf, nbytes) != (ssize_t)nbytes)
286 r = 1;
287 else {
288 while ((r = sread(fd, buf, sizeof(buf), 1)) > 0)
289 if (swrite(tfd, buf, (size_t)r) != r)
290 break;
291 }
292
293 switch (r) {
294 case -1:
295 file_error(ms, errno, "error copying from pipe to temp file");
296 return -1;
297 case 0:
298 break;
299 default:
300 file_error(ms, errno, "error while writing to temp file");
301 return -1;
302 }
303
304 /*
305 * We duplicate the file descriptor, because fclose on a
306 * tmpfile will delete the file, but any open descriptors
307 * can still access the phantom inode.
308 */
309 if ((fd = dup2(tfd, fd)) == -1) {
310 file_error(ms, errno, "could not dup descriptor for temp file");
311 return -1;
312 }
313 (void)close(tfd);
314 if (FINFO_LSEEK_FUNC(fd, (zend_off_t)0, SEEK_SET) == (zend_off_t)-1) {
315 file_badseek(ms);
316 return -1;
317 }
318 return fd;
319 }
320
321 #ifdef PHP_FILEINFO_UNCOMPRESS
322 #ifdef BUILTIN_DECOMPRESS
323
324 #define FHCRC (1 << 1)
325 #define FEXTRA (1 << 2)
326 #define FNAME (1 << 3)
327 #define FCOMMENT (1 << 4)
328
329 private size_t
uncompressgzipped(struct magic_set * ms,const unsigned char * old,unsigned char ** newch,size_t n)330 uncompressgzipped(struct magic_set *ms, const unsigned char *old,
331 unsigned char **newch, size_t n)
332 {
333 unsigned char flg = old[3];
334 size_t data_start = 10;
335 z_stream z;
336 int rc;
337
338 if (flg & FEXTRA) {
339 if (data_start+1 >= n)
340 return 0;
341 data_start += 2 + old[data_start] + old[data_start + 1] * 256;
342 }
343 if (flg & FNAME) {
344 while(data_start < n && old[data_start])
345 data_start++;
346 data_start++;
347 }
348 if(flg & FCOMMENT) {
349 while(data_start < n && old[data_start])
350 data_start++;
351 data_start++;
352 }
353 if(flg & FHCRC)
354 data_start += 2;
355
356 if (data_start >= n)
357 return 0;
358 if ((*newch = CAST(unsigned char *, emalloc(HOWMANY + 1))) == NULL) {
359 return 0;
360 }
361
362 /* XXX: const castaway, via strchr */
363 z.next_in = (Bytef *)strchr((const char *)old + data_start,
364 old[data_start]);
365 z.avail_in = CAST(uint32_t, (n - data_start));
366 z.next_out = *newch;
367 z.avail_out = HOWMANY;
368 z.zalloc = Z_NULL;
369 z.zfree = Z_NULL;
370 z.opaque = Z_NULL;
371
372 /* LINTED bug in header macro */
373 rc = inflateInit2(&z, -15);
374 if (rc != Z_OK) {
375 file_error(ms, 0, "zlib: %s", z.msg);
376 return 0;
377 }
378
379 rc = inflate(&z, Z_SYNC_FLUSH);
380 if (rc != Z_OK && rc != Z_STREAM_END) {
381 file_error(ms, 0, "zlib: %s", z.msg);
382 return 0;
383 }
384
385 n = (size_t)z.total_out;
386 (void)inflateEnd(&z);
387
388 /* let's keep the nul-terminate tradition */
389 (*newch)[n] = '\0';
390
391 return n;
392 }
393 #endif
394
395 private size_t
uncompressbuf(struct magic_set * ms,int fd,size_t method,const unsigned char * old,unsigned char ** newch,size_t n)396 uncompressbuf(struct magic_set *ms, int fd, size_t method,
397 const unsigned char *old, unsigned char **newch, size_t n)
398 {
399 int fdin[2], fdout[2];
400 int status;
401 ssize_t r;
402
403 #ifdef BUILTIN_DECOMPRESS
404 /* FIXME: This doesn't cope with bzip2 */
405 if (method == 2)
406 return uncompressgzipped(ms, old, newch, n);
407 #endif
408 (void)fflush(stdout);
409 (void)fflush(stderr);
410
411 if ((fd != -1 && pipe(fdin) == -1) || pipe(fdout) == -1) {
412 file_error(ms, errno, "cannot create pipe");
413 return NODATA;
414 }
415 switch (fork()) {
416 case 0: /* child */
417 (void) close(0);
418 if (fd != -1) {
419 (void) dup(fd);
420 (void) FINFO_LSEEK_FUNC(0, (zend_off_t)0, SEEK_SET);
421 } else {
422 (void) dup(fdin[0]);
423 (void) close(fdin[0]);
424 (void) close(fdin[1]);
425 }
426
427 (void) close(1);
428 (void) dup(fdout[1]);
429 (void) close(fdout[0]);
430 (void) close(fdout[1]);
431 #ifndef DEBUG
432 if (compr[method].silent)
433 (void)close(2);
434 #endif
435
436 (void)execvp(compr[method].argv[0],
437 (char *const *)(intptr_t)compr[method].argv);
438 #ifdef DEBUG
439 (void)fprintf(stderr, "exec `%s' failed (%s)\n",
440 compr[method].argv[0], strerror(errno));
441 #endif
442 exit(1);
443 /*NOTREACHED*/
444 case -1:
445 file_error(ms, errno, "could not fork");
446 return NODATA;
447
448 default: /* parent */
449 (void) close(fdout[1]);
450 if (fd == -1) {
451 (void) close(fdin[0]);
452 /*
453 * fork again, to avoid blocking because both
454 * pipes filled
455 */
456 switch (fork()) {
457 case 0: /* child */
458 (void)close(fdout[0]);
459 if (swrite(fdin[1], old, n) != (ssize_t)n) {
460 #ifdef DEBUG
461 (void)fprintf(stderr,
462 "Write failed (%s)\n",
463 strerror(errno));
464 #endif
465 exit(1);
466 }
467 exit(0);
468 /*NOTREACHED*/
469
470 case -1:
471 #ifdef DEBUG
472 (void)fprintf(stderr, "Fork failed (%s)\n",
473 strerror(errno));
474 #endif
475 exit(1);
476 /*NOTREACHED*/
477
478 default: /* parent */
479 break;
480 }
481 (void) close(fdin[1]);
482 fdin[1] = -1;
483 }
484
485 *newch = (unsigned char *) emalloc(HOWMANY + 1);
486
487 if ((r = sread(fdout[0], *newch, HOWMANY, 0)) <= 0) {
488 #ifdef DEBUG
489 (void)fprintf(stderr, "Read failed (%s)\n",
490 strerror(errno));
491 #endif
492 efree(*newch);
493 n = NODATA-;
494 *newch = NULL;
495 goto err;
496 } else {
497 n = r;
498 }
499 /* NUL terminate, as every buffer is handled here. */
500 (*newch)[n] = '\0';
501 err:
502 if (fdin[1] != -1)
503 (void) close(fdin[1]);
504 (void) close(fdout[0]);
505 #ifdef WNOHANG
506 while (waitpid(pid, NULL, WNOHANG) != -1)
507 continue;
508 #else
509 (void)wait(NULL);
510 #endif
511
512 (void) close(fdin[0]);
513
514 return n;
515 }
516 }
517 #endif /* if PHP_FILEINFO_UNCOMPRESS */
518
