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 * file - find type of a file or files - main program.
30 */
31
32 #include "file.h"
33
34 #ifndef lint
35 FILE_RCSID("@(#)$File: file.c,v 1.136 2009/12/06 23:18:04 rrt Exp $")
36 #endif /* lint */
37
38 #include "magic.h"
39
40 #include <stdlib.h>
41 #include <unistd.h>
42 #include <string.h>
43 #ifdef RESTORE_TIME
44 # if (__COHERENT__ >= 0x420)
45 # include <sys/utime.h>
46 # else
47 # ifdef USE_UTIMES
48 # include <sys/time.h>
49 # else
50 # include <utime.h>
51 # endif
52 # endif
53 #endif
54 #ifdef HAVE_UNISTD_H
55 #include <unistd.h> /* for read() */
56 #endif
57 #ifdef HAVE_LOCALE_H
58 #include <locale.h>
59 #endif
60 #ifdef HAVE_WCHAR_H
61 #include <wchar.h>
62 #endif
63
64 #if defined(HAVE_GETOPT_H) && defined(HAVE_STRUCT_OPTION)
65 #include <getopt.h>
66 #ifndef HAVE_GETOPT_LONG
67 int getopt_long(int argc, char * const *argv, const char *optstring, const struct option *longopts, int *longindex);
68 #endif
69 #else
70 #include "mygetopt.h"
71 #endif
72
73 #include "patchlevel.h"
74
75 #ifdef S_IFLNK
76 #define FILE_FLAGS "-bchikLNnprsvz0"
77 #else
78 #define FILE_FLAGS "-bcikNnprsvz0"
79 #endif
80
81 # define USAGE \
82 "Usage: %s [" FILE_FLAGS \
83 "] [--apple] [--mime-encoding] [--mime-type]\n" \
84 " [-e testname] [-F separator] [-f namefile] [-m magicfiles] file ...\n" \
85 " %s -C [-m magicfiles]\n" \
86 " %s [--help]\n"
87
88 #ifndef MAXPATHLEN
89 #define MAXPATHLEN 1024
90 #endif
91
92 private int /* Global command-line options */
93 bflag = 0, /* brief output format */
94 nopad = 0, /* Don't pad output */
95 nobuffer = 0, /* Do not buffer stdout */
96 nulsep = 0; /* Append '\0' to the separator */
97
98 private const char *separator = ":"; /* Default field separator */
99 private const struct option long_options[] = {
100 #define OPT(shortname, longname, opt, doc) \
101 {longname, opt, NULL, shortname},
102 #define OPT_LONGONLY(longname, opt, doc) \
103 {longname, opt, NULL, 0},
104 #include "file_opts.h"
105 #undef OPT
106 #undef OPT_LONGONLY
107 {0, 0, NULL, 0}
108 };
109 #define OPTSTRING "bcCde:f:F:hikLm:nNprsvz0"
110
111 private const struct {
112 const char *name;
113 int value;
114 } nv[] = {
115 { "apptype", MAGIC_NO_CHECK_APPTYPE },
116 { "ascii", MAGIC_NO_CHECK_ASCII },
117 { "cdf", MAGIC_NO_CHECK_CDF },
118 { "compress", MAGIC_NO_CHECK_COMPRESS },
119 { "elf", MAGIC_NO_CHECK_ELF },
120 { "encoding", MAGIC_NO_CHECK_ENCODING },
121 { "soft", MAGIC_NO_CHECK_SOFT },
122 { "tar", MAGIC_NO_CHECK_TAR },
123 { "tokens", MAGIC_NO_CHECK_TOKENS },
124 };
125
126 private char *progname; /* used throughout */
127
128 private void usage(void);
129 private void help(void);
130 int main(int, char *[]);
131
132 private int unwrap(struct magic_set *, const char *);
133 private int process(struct magic_set *ms, const char *, int);
134 private struct magic_set *load(const char *, int);
135
136
137 /*
138 * main - parse arguments and handle options
139 */
140 int
main(int argc,char * argv[])141 main(int argc, char *argv[])
142 {
143 int c;
144 size_t i;
145 int action = 0, didsomefiles = 0, errflg = 0;
146 int flags = 0, e = 0;
147 struct magic_set *magic = NULL;
148 int longindex;
149 const char *magicfile = NULL; /* where the magic is */
150
151 /* makes islower etc work for other langs */
152 (void)setlocale(LC_CTYPE, "");
153
154 #ifdef __EMX__
155 /* sh-like wildcard expansion! Shouldn't hurt at least ... */
156 _wildcard(&argc, &argv);
157 #endif
158
159 if ((progname = strrchr(argv[0], '/')) != NULL)
160 progname++;
161 else
162 progname = argv[0];
163
164 #ifdef S_IFLNK
165 flags |= getenv("POSIXLY_CORRECT") ? MAGIC_SYMLINK : 0;
166 #endif
167 while ((c = getopt_long(argc, argv, OPTSTRING, long_options,
168 &longindex)) != -1)
169 switch (c) {
170 case 0 :
171 switch (longindex) {
172 case 0:
173 help();
174 break;
175 case 10:
176 flags |= MAGIC_APPLE;
177 break;
178 case 11:
179 flags |= MAGIC_MIME_TYPE;
180 break;
181 case 12:
182 flags |= MAGIC_MIME_ENCODING;
183 break;
184 }
185 break;
186 case '0':
187 nulsep = 1;
188 break;
189 case 'b':
190 bflag++;
191 break;
192 case 'c':
193 action = FILE_CHECK;
194 break;
195 case 'C':
196 action = FILE_COMPILE;
197 break;
198 case 'd':
199 flags |= MAGIC_DEBUG|MAGIC_CHECK;
200 break;
201 case 'e':
202 for (i = 0; i < sizeof(nv) / sizeof(nv[0]); i++)
203 if (strcmp(nv[i].name, optarg) == 0)
204 break;
205
206 if (i == sizeof(nv) / sizeof(nv[0]))
207 errflg++;
208 else
209 flags |= nv[i].value;
210 break;
211
212 case 'f':
213 if(action)
214 usage();
215 if (magic == NULL)
216 if ((magic = load(magicfile, flags)) == NULL)
217 return 1;
218 e |= unwrap(magic, optarg);
219 ++didsomefiles;
220 break;
221 case 'F':
222 separator = optarg;
223 break;
224 case 'i':
225 flags |= MAGIC_MIME;
226 break;
227 case 'k':
228 flags |= MAGIC_CONTINUE;
229 break;
230 case 'm':
231 magicfile = optarg;
232 break;
233 case 'n':
234 ++nobuffer;
235 break;
236 case 'N':
237 ++nopad;
238 break;
239 #if defined(HAVE_UTIME) || defined(HAVE_UTIMES)
240 case 'p':
241 flags |= MAGIC_PRESERVE_ATIME;
242 break;
243 #endif
244 case 'r':
245 flags |= MAGIC_RAW;
246 break;
247 case 's':
248 flags |= MAGIC_DEVICES;
249 break;
250 case 'v':
251 if (magicfile == NULL)
252 magicfile = magic_getpath(magicfile, action);
253 (void)fprintf(stderr, "%s-%d.%.2d\n", progname,
254 FILE_VERSION_MAJOR, patchlevel);
255 (void)fprintf(stderr, "magic file from %s\n",
256 magicfile);
257 return 1;
258 case 'z':
259 flags |= MAGIC_COMPRESS;
260 break;
261 #ifdef S_IFLNK
262 case 'L':
263 flags |= MAGIC_SYMLINK;
264 break;
265 case 'h':
266 flags &= ~MAGIC_SYMLINK;
267 break;
268 #endif
269 case '?':
270 default:
271 errflg++;
272 break;
273 }
274
275 if (errflg) {
276 usage();
277 }
278 if (e)
279 return e;
280
281 switch(action) {
282 case FILE_CHECK:
283 case FILE_COMPILE:
284 /*
285 * Don't try to check/compile ~/.magic unless we explicitly
286 * ask for it.
287 */
288 magic = magic_open(flags|MAGIC_CHECK);
289 if (magic == NULL) {
290 (void)fprintf(stderr, "%s: %s\n", progname,
291 strerror(errno));
292 return 1;
293 }
294 c = action == FILE_CHECK ? magic_check(magic, magicfile) :
295 magic_compile(magic, magicfile);
296 if (c == -1) {
297 (void)fprintf(stderr, "%s: %s\n", progname,
298 magic_error(magic));
299 return 1;
300 }
301 return 0;
302 default:
303 if (magic == NULL)
304 if ((magic = load(magicfile, flags)) == NULL)
305 return 1;
306 break;
307 }
308
309 if (optind == argc) {
310 if (!didsomefiles)
311 usage();
312 }
313 else {
314 size_t j, wid, nw;
315 for (wid = 0, j = (size_t)optind; j < (size_t)argc; j++) {
316 nw = file_mbswidth(argv[j]);
317 if (nw > wid)
318 wid = nw;
319 }
320 /*
321 * If bflag is only set twice, set it depending on
322 * number of files [this is undocumented, and subject to change]
323 */
324 if (bflag == 2) {
325 bflag = optind >= argc - 1;
326 }
327 for (; optind < argc; optind++)
328 e |= process(magic, argv[optind], wid);
329 }
330
331 if (magic)
332 magic_close(magic);
333 return e;
334 }
335
336
337 private struct magic_set *
338 /*ARGSUSED*/
load(const char * magicfile,int flags)339 load(const char *magicfile, int flags)
340 {
341 struct magic_set *magic = magic_open(flags);
342 if (magic == NULL) {
343 (void)fprintf(stderr, "%s: %s\n", progname, strerror(errno));
344 return NULL;
345 }
346 if (magic_load(magic, magicfile) == -1) {
347 (void)fprintf(stderr, "%s: %s\n",
348 progname, magic_error(magic));
349 magic_close(magic);
350 return NULL;
351 }
352 return magic;
353 }
354
355 /*
356 * unwrap -- read a file of filenames, do each one.
357 */
358 private int
unwrap(struct magic_set * ms,const char * fn)359 unwrap(struct magic_set *ms, const char *fn)
360 {
361 char buf[MAXPATHLEN];
362 FILE *f;
363 int wid = 0, cwid;
364 int e = 0;
365
366 if (strcmp("-", fn) == 0) {
367 f = stdin;
368 wid = 1;
369 } else {
370 if ((f = fopen(fn, "r")) == NULL) {
371 (void)fprintf(stderr, "%s: Cannot open `%s' (%s).\n",
372 progname, fn, strerror(errno));
373 return 1;
374 }
375
376 while (fgets(buf, sizeof(buf), f) != NULL) {
377 buf[strcspn(buf, "\n")] = '\0';
378 cwid = file_mbswidth(buf);
379 if (cwid > wid)
380 wid = cwid;
381 }
382
383 rewind(f);
384 }
385
386 while (fgets(buf, sizeof(buf), f) != NULL) {
387 buf[strcspn(buf, "\n")] = '\0';
388 e |= process(ms, buf, wid);
389 if(nobuffer)
390 (void)fflush(stdout);
391 }
392
393 (void)fclose(f);
394 return e;
395 }
396
397 /*
398 * Called for each input file on the command line (or in a list of files)
399 */
400 private int
process(struct magic_set * ms,const char * inname,int wid)401 process(struct magic_set *ms, const char *inname, int wid)
402 {
403 const char *type;
404 int std_in = strcmp(inname, "-") == 0;
405
406 if (wid > 0 && !bflag) {
407 (void)printf("%s", std_in ? "/dev/stdin" : inname);
408 if (nulsep)
409 (void)putc('\0', stdout);
410 else
411 (void)printf("%s", separator);
412 (void)printf("%*s ",
413 (int) (nopad ? 0 : (wid - file_mbswidth(inname))), "");
414 }
415
416 type = magic_file(ms, std_in ? NULL : inname);
417 if (type == NULL) {
418 (void)printf("ERROR: %s\n", magic_error(ms));
419 return 1;
420 } else {
421 (void)printf("%s\n", type);
422 return 0;
423 }
424 }
425
426 size_t
file_mbswidth(const char * s)427 file_mbswidth(const char *s)
428 {
429 #if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH)
430 size_t bytesconsumed, old_n, n, width = 0;
431 mbstate_t state;
432 wchar_t nextchar;
433 (void)memset(&state, 0, sizeof(mbstate_t));
434 old_n = n = strlen(s);
435
436 while (n > 0) {
437 bytesconsumed = mbrtowc(&nextchar, s, n, &state);
438 if (bytesconsumed == (size_t)(-1) ||
439 bytesconsumed == (size_t)(-2)) {
440 /* Something went wrong, return something reasonable */
441 return old_n;
442 }
443 if (s[0] == '\n') {
444 /*
445 * do what strlen() would do, so that caller
446 * is always right
447 */
448 width++;
449 } else
450 width += wcwidth(nextchar);
451
452 s += bytesconsumed, n -= bytesconsumed;
453 }
454 return width;
455 #else
456 return strlen(s);
457 #endif
458 }
459
460 private void
usage(void)461 usage(void)
462 {
463 (void)fprintf(stderr, USAGE, progname, progname, progname);
464 exit(1);
465 }
466
467 private void
help(void)468 help(void)
469 {
470 (void)fputs(
471 "Usage: file [OPTION...] [FILE...]\n"
472 "Determine type of FILEs.\n"
473 "\n", stderr);
474 #define OPT(shortname, longname, opt, doc) \
475 fprintf(stderr, " -%c, --" longname doc, shortname);
476 #define OPT_LONGONLY(longname, opt, doc) \
477 fprintf(stderr, " --" longname doc);
478 #include "file_opts.h"
479 #undef OPT
480 #undef OPT_LONGONLY
481 exit(0);
482 }
483