2 * Copyright (c) Christos Zoulas 2003.
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions
8 * 1. Redistributions of source code must retain the above copyright
9 * notice immediately at the beginning of the file, without modification,
10 * this list of conditions, and the following disclaimer.
11 * 2. Redistributions in binary form must reproduce the above copyright
12 * notice, this list of conditions and the following disclaimer in the
13 * documentation and/or other materials provided with the distribution.
15 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
16 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
17 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
18 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE FOR
19 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
20 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
21 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
22 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
23 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
24 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
36 FILE_RCSID("@(#)$File: magic.c,v 1.117 2021/12/06 15:33:00 christos Exp $")
47 #include <limits.h> /* for PIPE_BUF */
49 #if defined(HAVE_UTIMES)
50 # include <sys/time.h>
51 #elif defined(HAVE_UTIME)
52 # if defined(HAVE_SYS_UTIME_H)
53 # include <sys/utime.h>
54 # elif defined(HAVE_UTIME_H)
60 #include <unistd.h> /* for read() */
64 /* Get the PIPE_BUF from pathconf */
66 #define PIPE_BUF pathconf(".", _PC_PIPE_BUF)
72 private void close_and_restore(const struct magic_set *, const char *, int,
74 private int unreadable_info(struct magic_set *, mode_t, const char *);
75 private const char* get_default_magic(void);
77 private const char *file_or_fd(struct magic_set *, const char *, int);
81 #define STDIN_FILENO 0
85 /* HINSTANCE of this shared library. Needed for get_default_magic() */
86 static HINSTANCE _w32_dll_instance = NULL;
89 _w32_append_path(char **hmagicpath, const char *fmt, ...)
96 if (vasprintf(&tmppath, fmt, ap) < 0) {
102 if (access(tmppath, R_OK) == -1)
105 if (*hmagicpath == NULL) {
106 *hmagicpath = tmppath;
110 if (asprintf(&newpath, "%s%c%s", *hmagicpath, PATHSEP, tmppath) < 0)
115 *hmagicpath = newpath;
122 _w32_get_magic_relative_to(char **hmagicpath, HINSTANCE module)
124 static const char *trypaths[] = {
125 "%s/share/misc/magic.mgc",
131 dllpath = calloc(MAX_PATH + 1, sizeof(*dllpath));
133 if (!GetModuleFileNameA(module, dllpath, MAX_PATH))
136 PathRemoveFileSpecA(dllpath);
139 char exepath[MAX_PATH];
140 GetModuleFileNameA(NULL, exepath, MAX_PATH);
141 PathRemoveFileSpecA(exepath);
142 if (stricmp(exepath, dllpath) == 0)
146 sp = strlen(dllpath);
147 if (sp > 3 && stricmp(&dllpath[sp - 3], "bin") == 0) {
148 _w32_append_path(hmagicpath,
149 "%s/../share/misc/magic.mgc", dllpath);
153 for (sp = 0; sp < __arraycount(trypaths); sp++)
154 _w32_append_path(hmagicpath, trypaths[sp], dllpath);
159 #ifndef BUILD_AS_WINDOWS_STATIC_LIBARAY
160 /* Placate GCC by offering a sacrificial previous prototype */
161 BOOL WINAPI DllMain(HINSTANCE, DWORD, LPVOID);
164 DllMain(HINSTANCE hinstDLL, DWORD fdwReason,
165 LPVOID lpvReserved __attribute__((__unused__)))
167 if (fdwReason == DLL_PROCESS_ATTACH)
168 _w32_dll_instance = hinstDLL;
175 get_default_magic(void)
177 static const char hmagic[] = "/.magic/magic.mgc";
178 static char *default_magic;
179 char *home, *hmagicpath;
186 default_magic = NULL;
188 if ((home = getenv("HOME")) == NULL)
191 if (asprintf(&hmagicpath, "%s/.magic.mgc", home) < 0)
193 if (stat(hmagicpath, &st) == -1) {
195 if (asprintf(&hmagicpath, "%s/.magic", home) < 0)
197 if (stat(hmagicpath, &st) == -1)
199 if (S_ISDIR(st.st_mode)) {
201 if (asprintf(&hmagicpath, "%s/%s", home, hmagic) < 0)
203 if (access(hmagicpath, R_OK) == -1)
208 if (asprintf(&default_magic, "%s:%s", hmagicpath, MAGIC) < 0)
211 return default_magic;
213 default_magic = NULL;
221 default_magic = NULL;
224 /* Before anything else, try to get a magic file from user HOME */
225 if ((home = getenv("HOME")) != NULL)
226 _w32_append_path(&hmagicpath, "%s%s", home, hmagic);
228 /* First, try to get a magic file from user-application data */
229 if ((home = getenv("LOCALAPPDATA")) != NULL)
230 _w32_append_path(&hmagicpath, "%s%s", home, hmagic);
232 /* Second, try to get a magic file from the user profile data */
233 if ((home = getenv("USERPROFILE")) != NULL)
234 _w32_append_path(&hmagicpath,
235 "%s/Local Settings/Application Data%s", home, hmagic);
237 /* Third, try to get a magic file from Common Files */
238 if ((home = getenv("COMMONPROGRAMFILES")) != NULL)
239 _w32_append_path(&hmagicpath, "%s%s", home, hmagic);
241 /* Fourth, try to get magic file relative to exe location */
242 _w32_get_magic_relative_to(&hmagicpath, NULL);
244 /* Fifth, try to get magic file relative to dll location */
245 _w32_get_magic_relative_to(&hmagicpath, _w32_dll_instance);
247 /* Avoid MAGIC constant - it likely points to a file within MSys tree */
248 default_magic = hmagicpath;
249 return default_magic;
254 magic_getpath(const char *magicfile, int action)
256 if (magicfile != NULL)
259 magicfile = getenv("MAGIC");
260 if (magicfile != NULL)
263 return action == FILE_LOAD ? get_default_magic() : MAGIC;
266 public struct magic_set *
267 magic_open(int flags)
269 return file_ms_alloc(flags);
273 unreadable_info(struct magic_set *ms, mode_t md, const char *file)
276 /* We cannot open it, but we were able to stat it. */
277 if (access(file, W_OK) == 0)
278 if (file_printf(ms, "writable, ") == -1)
281 if (access(file, X_OK) == 0)
282 if (file_printf(ms, "executable, ") == -1)
285 /* X_OK doesn't work well on MS-Windows */
287 const char *p = strrchr(file, '.');
288 if (p && (stricmp(p, ".exe")
289 || stricmp(p, ".dll")
290 || stricmp(p, ".bat")
291 || stricmp(p, ".cmd")))
292 if (file_printf(ms, "writable, ") == -1)
298 if (file_printf(ms, "regular file, ") == -1)
300 if (file_printf(ms, "no read permission") == -1)
306 magic_close(struct magic_set *ms)
317 magic_load(struct magic_set *ms, const char *magicfile)
321 return file_apprentice(ms, magicfile, FILE_LOAD);
326 * Install a set of compiled magic buffers.
329 magic_load_buffers(struct magic_set *ms, void **bufs, size_t *sizes,
334 return buffer_apprentice(ms, RCAST(struct magic **, bufs),
340 magic_compile(struct magic_set *ms, const char *magicfile)
344 return file_apprentice(ms, magicfile, FILE_COMPILE);
348 magic_check(struct magic_set *ms, const char *magicfile)
352 return file_apprentice(ms, magicfile, FILE_CHECK);
356 magic_list(struct magic_set *ms, const char *magicfile)
360 return file_apprentice(ms, magicfile, FILE_LIST);
364 close_and_restore(const struct magic_set *ms, const char *name, int fd,
365 const struct stat *sb)
367 if (fd == STDIN_FILENO || name == NULL)
371 if ((ms->flags & MAGIC_PRESERVE_ATIME) != 0) {
373 * Try to restore access, modification times if read it.
374 * This is really *bad* because it will modify the status
375 * time of the file... And of course this will affect
379 struct timeval utsbuf[2];
380 (void)memset(utsbuf, 0, sizeof(utsbuf));
381 utsbuf[0].tv_sec = sb->st_atime;
382 utsbuf[1].tv_sec = sb->st_mtime;
384 (void) utimes(name, utsbuf); /* don't care if loses */
385 #elif defined(HAVE_UTIME_H) || defined(HAVE_SYS_UTIME_H)
386 struct utimbuf utbuf;
388 (void)memset(&utbuf, 0, sizeof(utbuf));
389 utbuf.actime = sb->st_atime;
390 utbuf.modtime = sb->st_mtime;
391 (void) utime(name, &utbuf); /* don't care if loses */
399 * find type of descriptor
402 magic_descriptor(struct magic_set *ms, int fd)
406 return file_or_fd(ms, NULL, fd);
410 * find type of named file
413 magic_file(struct magic_set *ms, const char *inname)
417 return file_or_fd(ms, inname, STDIN_FILENO);
421 file_or_fd(struct magic_set *ms, const char *inname, int fd)
426 ssize_t nbytes = 0; /* number of bytes read from a datafile */
429 off_t pos = CAST(off_t, -1);
431 if (file_reset(ms, 1) == -1)
435 * one extra for terminating '\0', and
436 * some overlapping space for matches near EOF
438 #define SLOP (1 + sizeof(union VALUETYPE))
439 if ((buf = CAST(unsigned char *, malloc(ms->bytes_max + SLOP))) == NULL)
442 switch (file_fsmagic(ms, inname, &sb)) {
445 case 0: /* nothing found */
447 default: /* matched it and printed type */
453 /* Place stdin in binary mode, so EOF (Ctrl+Z) doesn't stop early. */
454 if (fd == STDIN_FILENO)
455 _setmode(STDIN_FILENO, O_BINARY);
457 if (inname != NULL) {
458 int flags = O_RDONLY|O_BINARY|O_NONBLOCK|O_CLOEXEC;
460 if ((fd = open(inname, flags)) < 0) {
461 okstat = stat(inname, &sb) == 0;
464 * Can't stat, can't open. It may have been opened in
465 * fsmagic, so if the user doesn't have read permission,
466 * allow it to say so; otherwise an error was probably
467 * displayed in fsmagic.
469 if (!okstat && errno == EACCES) {
470 sb.st_mode = S_IFBLK;
475 unreadable_info(ms, sb.st_mode, inname) == -1)
480 #if O_CLOEXEC == 0 && defined(F_SETFD)
481 (void)fcntl(fd, F_SETFD, FD_CLOEXEC);
486 okstat = fstat(fd, &sb) == 0;
487 if (okstat && S_ISFIFO(sb.st_mode))
490 pos = lseek(fd, CAST(off_t, 0), SEEK_CUR);
494 * try looking at the first ms->bytes_max bytes
500 while ((r = sread(fd, RCAST(void *, &buf[nbytes]),
501 CAST(size_t, ms->bytes_max - nbytes), 1)) > 0) {
503 if (r < PIPE_BUF) break;
507 if (nbytes == 0 && inname) {
508 /* We can not read it, but we were able to stat it. */
509 if (unreadable_info(ms, sb.st_mode, inname) == -1)
515 } else if (fd != -1) {
516 /* Windows refuses to read from a big console buffer. */
519 _isatty(fd) ? 8 * 1024 :
522 if ((nbytes = read(fd, RCAST(void *, buf), howmany)) == -1) {
523 if (inname == NULL && fd != STDIN_FILENO)
524 file_error(ms, errno, "cannot read fd %d", fd);
526 file_error(ms, errno, "cannot read `%s'",
527 inname == NULL ? "/dev/stdin" : inname);
532 (void)memset(buf + nbytes, 0, SLOP); /* NUL terminate */
533 if (file_buffer(ms, fd, okstat ? &sb : NULL, inname, buf, CAST(size_t, nbytes)) == -1)
539 if (pos != CAST(off_t, -1))
540 (void)lseek(fd, pos, SEEK_SET);
541 close_and_restore(ms, inname, fd, &sb);
544 return rv == 0 ? file_getbuffer(ms) : NULL;
549 magic_buffer(struct magic_set *ms, const void *buf, size_t nb)
553 if (file_reset(ms, 1) == -1)
556 * The main work is done here!
557 * We have the file name and/or the data buffer to be identified.
559 if (file_buffer(ms, -1, NULL, NULL, buf, nb) == -1) {
562 return file_getbuffer(ms);
567 magic_error(struct magic_set *ms)
570 return "Magic database is not open";
571 return (ms->event_flags & EVENT_HAD_ERR) ? ms->o.buf : NULL;
575 magic_errno(struct magic_set *ms)
579 return (ms->event_flags & EVENT_HAD_ERR) ? ms->error : 0;
583 magic_getflags(struct magic_set *ms)
592 magic_setflags(struct magic_set *ms, int flags)
596 #if !defined(HAVE_UTIME) && !defined(HAVE_UTIMES)
597 if (flags & MAGIC_PRESERVE_ATIME)
607 return MAGIC_VERSION;
611 magic_setparam(struct magic_set *ms, int param, const void *val)
616 case MAGIC_PARAM_INDIR_MAX:
617 ms->indir_max = CAST(uint16_t, *CAST(const size_t *, val));
619 case MAGIC_PARAM_NAME_MAX:
620 ms->name_max = CAST(uint16_t, *CAST(const size_t *, val));
622 case MAGIC_PARAM_ELF_PHNUM_MAX:
623 ms->elf_phnum_max = CAST(uint16_t, *CAST(const size_t *, val));
625 case MAGIC_PARAM_ELF_SHNUM_MAX:
626 ms->elf_shnum_max = CAST(uint16_t, *CAST(const size_t *, val));
628 case MAGIC_PARAM_ELF_NOTES_MAX:
629 ms->elf_notes_max = CAST(uint16_t, *CAST(const size_t *, val));
631 case MAGIC_PARAM_REGEX_MAX:
632 ms->regex_max = CAST(uint16_t, *CAST(const size_t *, val));
634 case MAGIC_PARAM_BYTES_MAX:
635 ms->bytes_max = *CAST(const size_t *, val);
637 case MAGIC_PARAM_ENCODING_MAX:
638 ms->encoding_max = *CAST(const size_t *, val);
647 magic_getparam(struct magic_set *ms, int param, void *val)
652 case MAGIC_PARAM_INDIR_MAX:
653 *CAST(size_t *, val) = ms->indir_max;
655 case MAGIC_PARAM_NAME_MAX:
656 *CAST(size_t *, val) = ms->name_max;
658 case MAGIC_PARAM_ELF_PHNUM_MAX:
659 *CAST(size_t *, val) = ms->elf_phnum_max;
661 case MAGIC_PARAM_ELF_SHNUM_MAX:
662 *CAST(size_t *, val) = ms->elf_shnum_max;
664 case MAGIC_PARAM_ELF_NOTES_MAX:
665 *CAST(size_t *, val) = ms->elf_notes_max;
667 case MAGIC_PARAM_REGEX_MAX:
668 *CAST(size_t *, val) = ms->regex_max;
670 case MAGIC_PARAM_BYTES_MAX:
671 *CAST(size_t *, val) = ms->bytes_max;
673 case MAGIC_PARAM_ENCODING_MAX:
674 *CAST(size_t *, val) = ms->encoding_max;