Update to file-4.19.
[dragonfly.git] / contrib / file-4 / src / file.c
CommitLineData
ab0b56cc
JS
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#include "magic.h"
34
35#include <stdio.h>
36#include <stdlib.h>
37#include <unistd.h>
38#include <string.h>
39#include <sys/types.h>
40#include <sys/param.h> /* for MAXPATHLEN */
41#include <sys/stat.h>
ab0b56cc
JS
42#ifdef RESTORE_TIME
43# if (__COHERENT__ >= 0x420)
44# include <sys/utime.h>
45# else
46# ifdef USE_UTIMES
47# include <sys/time.h>
48# else
49# include <utime.h>
50# endif
51# endif
52#endif
53#ifdef HAVE_UNISTD_H
54#include <unistd.h> /* for read() */
55#endif
56#ifdef HAVE_LOCALE_H
57#include <locale.h>
58#endif
59#ifdef HAVE_WCHAR_H
60#include <wchar.h>
61#endif
62
63#ifdef HAVE_GETOPT_H
64#include <getopt.h> /* for long options (is this portable?)*/
65#else
66#undef HAVE_GETOPT_LONG
67#endif
68
69#include <netinet/in.h> /* for byte swapping */
70
71#include "patchlevel.h"
72
73#ifndef lint
5fc399ce 74FILE_RCSID("@(#)$Id: file.c,v 1.104 2006/11/25 17:28:54 christos Exp $")
ab0b56cc
JS
75#endif /* lint */
76
77
78#ifdef S_IFLNK
9b22a626 79#define SYMLINKFLAG "Lh"
ab0b56cc
JS
80#else
81#define SYMLINKFLAG ""
82#endif
83
5fc399ce 84# define USAGE "Usage: %s [-bcik" SYMLINKFLAG "nNrsvz0] [-f namefile] [-F separator] [-m magicfiles] file...\n %s -C -m magicfiles\n"
ab0b56cc
JS
85
86#ifndef MAXPATHLEN
87#define MAXPATHLEN 512
88#endif
89
90private int /* Global command-line options */
91 bflag = 0, /* brief output format */
92 nopad = 0, /* Don't pad output */
5fc399ce
PA
93 nobuffer = 0, /* Do not buffer stdout */
94 nulsep = 0; /* Append '\0' to the separator */
ab0b56cc
JS
95
96private const char *magicfile = 0; /* where the magic is */
97private const char *default_magicfile = MAGIC;
2be182fc 98private const char *separator = ":"; /* Default field separator */
ab0b56cc
JS
99
100private char *progname; /* used throughout */
101
102private struct magic_set *magic;
103
104private void unwrap(char *);
105private void usage(void);
106#ifdef HAVE_GETOPT_LONG
107private void help(void);
108#endif
109#if 0
110private int byteconv4(int, int, int);
111private short byteconv2(int, int, int);
112#endif
113
114int main(int, char *[]);
115private void process(const char *, int);
116private void load(const char *, int);
117
118
119/*
120 * main - parse arguments and handle options
121 */
122int
123main(int argc, char *argv[])
124{
125 int c;
126 int action = 0, didsomefiles = 0, errflg = 0;
127 int flags = 0;
128 char *home, *usermagic;
129 struct stat sb;
9b22a626 130 static const char hmagic[] = "/.magic";
5fc399ce 131#define OPTSTRING "bcCdf:F:hikLm:nNprsvz0"
ab0b56cc
JS
132#ifdef HAVE_GETOPT_LONG
133 int longindex;
9b22a626 134 static const struct option long_options[] =
ab0b56cc
JS
135 {
136 {"version", 0, 0, 'v'},
137 {"help", 0, 0, 0},
138 {"brief", 0, 0, 'b'},
139 {"checking-printout", 0, 0, 'c'},
140 {"debug", 0, 0, 'd'},
141 {"files-from", 1, 0, 'f'},
142 {"separator", 1, 0, 'F'},
143 {"mime", 0, 0, 'i'},
144 {"keep-going", 0, 0, 'k'},
145#ifdef S_IFLNK
146 {"dereference", 0, 0, 'L'},
9b22a626 147 {"no-dereference", 0, 0, 'h'},
ab0b56cc
JS
148#endif
149 {"magic-file", 1, 0, 'm'},
150#if defined(HAVE_UTIME) || defined(HAVE_UTIMES)
151 {"preserve-date", 0, 0, 'p'},
152#endif
153 {"uncompress", 0, 0, 'z'},
154 {"raw", 0, 0, 'r'},
155 {"no-buffer", 0, 0, 'n'},
156 {"no-pad", 0, 0, 'N'},
157 {"special-files", 0, 0, 's'},
158 {"compile", 0, 0, 'C'},
5fc399ce 159 {"print0", 0, 0, '0'},
ab0b56cc
JS
160 {0, 0, 0, 0},
161 };
162#endif
163
164#ifdef LC_CTYPE
9b22a626
PA
165 /* makes islower etc work for other langs */
166 (void)setlocale(LC_CTYPE, "");
ab0b56cc
JS
167#endif
168
169#ifdef __EMX__
170 /* sh-like wildcard expansion! Shouldn't hurt at least ... */
171 _wildcard(&argc, &argv);
172#endif
173
174 if ((progname = strrchr(argv[0], '/')) != NULL)
175 progname++;
176 else
177 progname = argv[0];
178
179 magicfile = default_magicfile;
180 if ((usermagic = getenv("MAGIC")) != NULL)
181 magicfile = usermagic;
182 else
183 if ((home = getenv("HOME")) != NULL) {
9b22a626
PA
184 if ((usermagic = malloc(strlen(home)
185 + sizeof(hmagic))) != NULL) {
ab0b56cc 186 (void)strcpy(usermagic, home);
9b22a626 187 (void)strcat(usermagic, hmagic);
ab0b56cc
JS
188 if (stat(usermagic, &sb)<0)
189 free(usermagic);
190 else
191 magicfile = usermagic;
192 }
193 }
194
9b22a626
PA
195#ifdef S_IFLNK
196 flags |= getenv("POSIXLY_CORRECT") ? MAGIC_SYMLINK : 0;
197#endif
ab0b56cc
JS
198#ifndef HAVE_GETOPT_LONG
199 while ((c = getopt(argc, argv, OPTSTRING)) != -1)
200#else
201 while ((c = getopt_long(argc, argv, OPTSTRING, long_options,
202 &longindex)) != -1)
203#endif
204 switch (c) {
205#ifdef HAVE_GETOPT_LONG
206 case 0 :
207 if (longindex == 1)
208 help();
209 break;
210#endif
5fc399ce
PA
211 case '0':
212 nulsep = 1;
213 break;
ab0b56cc
JS
214 case 'b':
215 ++bflag;
216 break;
217 case 'c':
218 action = FILE_CHECK;
219 break;
220 case 'C':
221 action = FILE_COMPILE;
222 break;
223 case 'd':
224 flags |= MAGIC_DEBUG|MAGIC_CHECK;
225 break;
226 case 'f':
227 if(action)
228 usage();
229 load(magicfile, flags);
230 unwrap(optarg);
231 ++didsomefiles;
232 break;
233 case 'F':
234 separator = optarg;
235 break;
236 case 'i':
237 flags |= MAGIC_MIME;
238 break;
239 case 'k':
240 flags |= MAGIC_CONTINUE;
241 break;
242 case 'm':
243 magicfile = optarg;
244 break;
245 case 'n':
246 ++nobuffer;
247 break;
248 case 'N':
249 ++nopad;
250 break;
251#if defined(HAVE_UTIME) || defined(HAVE_UTIMES)
252 case 'p':
253 flags |= MAGIC_PRESERVE_ATIME;
254 break;
255#endif
256 case 'r':
257 flags |= MAGIC_RAW;
258 break;
259 case 's':
260 flags |= MAGIC_DEVICES;
261 break;
262 case 'v':
9b22a626 263 (void)fprintf(stdout, "%s-%d.%.2d\n", progname,
ab0b56cc 264 FILE_VERSION_MAJOR, patchlevel);
9b22a626 265 (void)fprintf(stdout, "magic file from %s\n",
ab0b56cc
JS
266 magicfile);
267 return 1;
268 case 'z':
269 flags |= MAGIC_COMPRESS;
270 break;
271#ifdef S_IFLNK
272 case 'L':
273 flags |= MAGIC_SYMLINK;
274 break;
9b22a626
PA
275 case 'h':
276 flags &= ~MAGIC_SYMLINK;
277 break;
ab0b56cc
JS
278#endif
279 case '?':
280 default:
281 errflg++;
282 break;
283 }
284
285 if (errflg) {
286 usage();
287 }
288
289 switch(action) {
290 case FILE_CHECK:
291 case FILE_COMPILE:
292 magic = magic_open(flags|MAGIC_CHECK);
293 if (magic == NULL) {
294 (void)fprintf(stderr, "%s: %s\n", progname,
295 strerror(errno));
296 return 1;
297 }
298 c = action == FILE_CHECK ? magic_check(magic, magicfile) :
299 magic_compile(magic, magicfile);
300 if (c == -1) {
301 (void)fprintf(stderr, "%s: %s\n", progname,
302 magic_error(magic));
303 return -1;
304 }
305 return 0;
306 default:
307 load(magicfile, flags);
308 break;
309 }
310
311 if (optind == argc) {
312 if (!didsomefiles) {
313 usage();
314 }
315 }
316 else {
317 int i, wid, nw;
318 for (wid = 0, i = optind; i < argc; i++) {
319 nw = file_mbswidth(argv[i]);
320 if (nw > wid)
321 wid = nw;
322 }
323 for (; optind < argc; optind++)
324 process(argv[optind], wid);
325 }
326
327 magic_close(magic);
328 return 0;
329}
330
331
332private void
9b22a626 333/*ARGSUSED*/
ab0b56cc
JS
334load(const char *m, int flags)
335{
336 if (magic)
337 return;
338 magic = magic_open(flags);
339 if (magic == NULL) {
340 (void)fprintf(stderr, "%s: %s\n", progname, strerror(errno));
341 exit(1);
342 }
343 if (magic_load(magic, magicfile) == -1) {
344 (void)fprintf(stderr, "%s: %s\n",
345 progname, magic_error(magic));
346 exit(1);
347 }
348}
349
350/*
351 * unwrap -- read a file of filenames, do each one.
352 */
353private void
354unwrap(char *fn)
355{
356 char buf[MAXPATHLEN];
357 FILE *f;
358 int wid = 0, cwid;
9b22a626 359 size_t len;
ab0b56cc
JS
360
361 if (strcmp("-", fn) == 0) {
362 f = stdin;
363 wid = 1;
364 } else {
365 if ((f = fopen(fn, "r")) == NULL) {
366 (void)fprintf(stderr, "%s: Cannot open `%s' (%s).\n",
367 progname, fn, strerror(errno));
368 exit(1);
369 }
370
371 while (fgets(buf, MAXPATHLEN, f) != NULL) {
9b22a626
PA
372 len = strlen(buf);
373 if (len > 0 && buf[len - 1] == '\n')
374 buf[len - 1] = '\0';
375 cwid = file_mbswidth(buf);
ab0b56cc
JS
376 if (cwid > wid)
377 wid = cwid;
378 }
379
380 rewind(f);
381 }
382
383 while (fgets(buf, MAXPATHLEN, f) != NULL) {
9b22a626
PA
384 len = strlen(buf);
385 if (len > 0 && buf[len - 1] == '\n')
386 buf[len - 1] = '\0';
ab0b56cc
JS
387 process(buf, wid);
388 if(nobuffer)
9b22a626 389 (void)fflush(stdout);
ab0b56cc
JS
390 }
391
9b22a626 392 (void)fclose(f);
ab0b56cc
JS
393}
394
9b22a626
PA
395/*
396 * Called for each input file on the command line (or in a list of files)
397 */
ab0b56cc
JS
398private void
399process(const char *inname, int wid)
400{
401 const char *type;
402 int std_in = strcmp(inname, "-") == 0;
403
5fc399ce
PA
404 if (wid > 0 && !bflag) {
405 (void)printf("%s", std_in ? "/dev/stdin" : inname);
406 if (nulsep)
407 (void)puts('\0');
408 else
409 (void)printf("%s", separator);
410 (void)printf("%*s ",
411 (int) (nopad ? 0 : (wid - file_mbswidth(inname))), "");
412 }
ab0b56cc
JS
413
414 type = magic_file(magic, std_in ? NULL : inname);
415 if (type == NULL)
9b22a626 416 (void)printf("ERROR: %s\n", magic_error(magic));
ab0b56cc 417 else
9b22a626 418 (void)printf("%s\n", type);
ab0b56cc
JS
419}
420
421
422#if 0
423/*
424 * byteconv4
425 * Input:
426 * from 4 byte quantity to convert
427 * same whether to perform byte swapping
428 * big_endian whether we are a big endian host
429 */
430private int
431byteconv4(int from, int same, int big_endian)
432{
433 if (same)
434 return from;
435 else if (big_endian) { /* lsb -> msb conversion on msb */
436 union {
437 int i;
438 char c[4];
439 } retval, tmpval;
440
441 tmpval.i = from;
442 retval.c[0] = tmpval.c[3];
443 retval.c[1] = tmpval.c[2];
444 retval.c[2] = tmpval.c[1];
445 retval.c[3] = tmpval.c[0];
446
447 return retval.i;
448 }
449 else
450 return ntohl(from); /* msb -> lsb conversion on lsb */
451}
452
453/*
454 * byteconv2
455 * Same as byteconv4, but for shorts
456 */
457private short
458byteconv2(int from, int same, int big_endian)
459{
460 if (same)
461 return from;
462 else if (big_endian) { /* lsb -> msb conversion on msb */
463 union {
464 short s;
465 char c[2];
466 } retval, tmpval;
467
468 tmpval.s = (short) from;
469 retval.c[0] = tmpval.c[1];
470 retval.c[1] = tmpval.c[0];
471
472 return retval.s;
473 }
474 else
475 return ntohs(from); /* msb -> lsb conversion on lsb */
476}
477#endif
478
479size_t
480file_mbswidth(const char *s)
481{
482#if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH)
483 size_t bytesconsumed, old_n, n, width = 0;
484 mbstate_t state;
485 wchar_t nextchar;
486 (void)memset(&state, 0, sizeof(mbstate_t));
487 old_n = n = strlen(s);
488
489 while (n > 0) {
490 bytesconsumed = mbrtowc(&nextchar, s, n, &state);
491 if (bytesconsumed == (size_t)(-1) ||
492 bytesconsumed == (size_t)(-2)) {
493 /* Something went wrong, return something reasonable */
494 return old_n;
495 }
496 if (s[0] == '\n') {
497 /*
498 * do what strlen() would do, so that caller
499 * is always right
500 */
501 width++;
502 } else
503 width += wcwidth(nextchar);
504
505 s += bytesconsumed, n -= bytesconsumed;
506 }
507 return width;
508#else
509 return strlen(s);
510#endif
511}
512
513private void
514usage(void)
515{
516 (void)fprintf(stderr, USAGE, progname, progname);
517#ifdef HAVE_GETOPT_LONG
518 (void)fputs("Try `file --help' for more information.\n", stderr);
519#endif
520 exit(1);
521}
522
523#ifdef HAVE_GETOPT_LONG
524private void
525help(void)
526{
9b22a626 527 (void)puts(
ab0b56cc
JS
528"Usage: file [OPTION]... [FILE]...\n"
529"Determine file type of FILEs.\n"
530"\n"
531" -m, --magic-file LIST use LIST as a colon-separated list of magic\n"
532" number files\n"
533" -z, --uncompress try to look inside compressed files\n"
534" -b, --brief do not prepend filenames to output lines\n"
535" -c, --checking-printout print the parsed form of the magic file, use in\n"
536" conjunction with -m to debug a new magic file\n"
537" before installing it\n"
538" -f, --files-from FILE read the filenames to be examined from FILE\n"
539" -F, --separator string use string as separator instead of `:'\n"
540" -i, --mime output mime type strings\n"
541" -k, --keep-going don't stop at the first match\n"
542" -L, --dereference causes symlinks to be followed\n"
543" -n, --no-buffer do not buffer output\n"
544" -N, --no-pad do not pad output\n"
545" -p, --preserve-date preserve access times on files\n"
546" -r, --raw don't translate unprintable chars to \\ooo\n"
547" -s, --special-files treat special (block/char devices) files as\n"
548" ordinary ones\n"
549" --help display this help and exit\n"
550" --version output version information and exit\n"
551);
552 exit(0);
553}
554#endif