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
30 FILE_RCSID("@(#)$File: funcs.c,v 1.104 2019/05/07 02:27:11 christos Exp $")
39 #if defined(HAVE_WCHAR_H)
42 #if defined(HAVE_WCTYPE_H)
48 #define SIZE_MAX ((size_t)~0)
52 * Like printf, only we append to a buffer.
55 file_vprintf(struct magic_set *ms, const char *fmt, va_list ap)
60 if (ms->event_flags & EVENT_HAD_ERR)
62 len = vasprintf(&buf, fmt, ap);
66 if (ms->o.buf != NULL) {
67 len = asprintf(&newstr, "%s%s", ms->o.buf, buf);
77 fprintf(stderr, "vasprintf failed (%s)", strerror(errno));
82 file_printf(struct magic_set *ms, const char *fmt, ...)
88 rv = file_vprintf(ms, fmt, ap);
94 * error - print best error message possible
97 __attribute__((__format__(__printf__, 3, 0)))
99 file_error_core(struct magic_set *ms, int error, const char *f, va_list va,
102 /* Only the first error is ok */
103 if (ms->event_flags & EVENT_HAD_ERR)
108 (void)file_printf(ms, "line %" SIZE_T_FORMAT "u:", lineno);
110 if (ms->o.buf && *ms->o.buf)
111 (void)file_printf(ms, " ");
112 (void)file_vprintf(ms, f, va);
114 (void)file_printf(ms, " (%s)", strerror(error));
115 ms->event_flags |= EVENT_HAD_ERR;
121 file_error(struct magic_set *ms, int error, const char *f, ...)
125 file_error_core(ms, error, f, va, 0);
130 * Print an error with magic line number.
134 file_magerror(struct magic_set *ms, const char *f, ...)
138 file_error_core(ms, 0, f, va, ms->line);
143 file_oomem(struct magic_set *ms, size_t len)
145 file_error(ms, errno, "cannot allocate %" SIZE_T_FORMAT "u bytes",
150 file_badseek(struct magic_set *ms)
152 file_error(ms, errno, "error seeking");
156 file_badread(struct magic_set *ms)
158 file_error(ms, errno, "error reading");
164 file_separator(struct magic_set *ms)
166 return file_printf(ms, "\n- ");
170 checkdone(struct magic_set *ms, int *rv)
172 if ((ms->flags & MAGIC_CONTINUE) == 0)
174 if (file_separator(ms) == -1)
180 file_default(struct magic_set *ms, size_t nb)
182 if (ms->flags & MAGIC_MIME) {
183 if ((ms->flags & MAGIC_MIME_TYPE) &&
184 file_printf(ms, "application/%s",
185 nb ? "octet-stream" : "x-empty") == -1)
189 if (ms->flags & MAGIC_APPLE) {
190 if (file_printf(ms, "UNKNUNKN") == -1)
194 if (ms->flags & MAGIC_EXTENSION) {
195 if (file_printf(ms, "???") == -1)
203 * The magic detection functions return:
210 file_buffer(struct magic_set *ms, int fd, struct stat *st,
211 const char *inname __attribute__ ((__unused__)),
212 const void *buf, size_t nb)
214 int m = 0, rv = 0, looks_text = 0;
215 const char *code = NULL;
216 const char *code_mime = "binary";
217 const char *def = "data";
218 const char *ftype = NULL;
222 buffer_init(&b, fd, st, buf, nb);
223 ms->mode = b.st.st_mode;
228 } else if (nb == 1) {
229 def = "very short file (no magic)";
233 if ((ms->flags & MAGIC_NO_CHECK_ENCODING) == 0) {
234 looks_text = file_encoding(ms, &b, NULL, 0,
235 &code, &code_mime, &ftype);
239 if ((ms->flags & MAGIC_NO_CHECK_APPTYPE) == 0 && inname) {
240 m = file_os2_apptype(ms, inname, &b);
241 if ((ms->flags & MAGIC_DEBUG) != 0)
242 (void)fprintf(stderr, "[try os2_apptype %d]\n", m);
254 /* try compression stuff */
255 if ((ms->flags & MAGIC_NO_CHECK_COMPRESS) == 0) {
256 m = file_zmagic(ms, &b, inname);
257 if ((ms->flags & MAGIC_DEBUG) != 0)
258 (void)fprintf(stderr, "[try zmagic %d]\n", m);
264 /* Check if we have a tar file */
265 if ((ms->flags & MAGIC_NO_CHECK_TAR) == 0) {
266 m = file_is_tar(ms, &b);
267 if ((ms->flags & MAGIC_DEBUG) != 0)
268 (void)fprintf(stderr, "[try tar %d]\n", m);
270 if (checkdone(ms, &rv))
275 /* Check if we have a JSON file */
276 if ((ms->flags & MAGIC_NO_CHECK_JSON) == 0) {
277 m = file_is_json(ms, &b);
278 if ((ms->flags & MAGIC_DEBUG) != 0)
279 (void)fprintf(stderr, "[try json %d]\n", m);
281 if (checkdone(ms, &rv))
286 /* Check if we have a CDF file */
287 if ((ms->flags & MAGIC_NO_CHECK_CDF) == 0) {
288 m = file_trycdf(ms, &b);
289 if ((ms->flags & MAGIC_DEBUG) != 0)
290 (void)fprintf(stderr, "[try cdf %d]\n", m);
292 if (checkdone(ms, &rv))
297 if ((ms->flags & MAGIC_NO_CHECK_ELF) == 0 && nb > 5 && fd != -1) {
300 * We matched something in the file, so this
301 * *might* be an ELF file, and the file is at
302 * least 5 bytes long, so if it's an ELF file
303 * it has at least one byte past the ELF magic
304 * number - try extracting information from the
305 * ELF headers that cannot easily be extracted
306 * with rules in the magic file. We we don't
307 * print the information yet.
309 if ((pb = file_push_buffer(ms)) == NULL)
312 rv = file_tryelf(ms, &b);
313 rbuf = file_pop_buffer(ms, pb);
318 if ((ms->flags & MAGIC_DEBUG) != 0)
319 (void)fprintf(stderr, "[try elf %d]\n", m);
323 /* try soft magic tests */
324 if ((ms->flags & MAGIC_NO_CHECK_SOFT) == 0) {
325 m = file_softmagic(ms, &b, NULL, NULL, BINTEST, looks_text);
326 if ((ms->flags & MAGIC_DEBUG) != 0)
327 (void)fprintf(stderr, "[try softmagic %d]\n", m);
328 if (m == 1 && rbuf) {
329 if (file_printf(ms, "%s", rbuf) == -1)
333 if (checkdone(ms, &rv))
338 /* try text properties */
339 if ((ms->flags & MAGIC_NO_CHECK_TEXT) == 0) {
341 m = file_ascmagic(ms, &b, looks_text);
342 if ((ms->flags & MAGIC_DEBUG) != 0)
343 (void)fprintf(stderr, "[try ascmagic %d]\n", m);
353 rv = file_default(ms, nb);
355 if (file_printf(ms, "%s", def) == -1)
359 if ((ms->flags & MAGIC_MIME_ENCODING) != 0) {
360 if (ms->flags & MAGIC_MIME_TYPE)
361 if (file_printf(ms, "; charset=") == -1)
363 if (file_printf(ms, "%s", code_mime) == -1)
379 file_reset(struct magic_set *ms, int checkloaded)
381 if (checkloaded && ms->mlist[0] == NULL) {
382 file_error(ms, 0, "no magic files loaded");
393 ms->event_flags &= ~EVENT_HAD_ERR;
398 #define OCTALIFY(n, o) \
400 (void)(*(n)++ = '\\', \
401 *(n)++ = ((CAST(uint32_t, *(o)) >> 6) & 3) + '0', \
402 *(n)++ = ((CAST(uint32_t, *(o)) >> 3) & 7) + '0', \
403 *(n)++ = ((CAST(uint32_t, *(o)) >> 0) & 7) + '0', \
406 protected const char *
407 file_getbuffer(struct magic_set *ms)
409 char *pbuf, *op, *np;
412 if (ms->event_flags & EVENT_HAD_ERR)
415 if (ms->flags & MAGIC_RAW)
418 if (ms->o.buf == NULL)
421 /* * 4 is for octal representation, + 1 is for NUL */
422 len = strlen(ms->o.buf);
423 if (len > (SIZE_MAX - 1) / 4) {
428 if ((pbuf = CAST(char *, realloc(ms->o.pbuf, psize))) == NULL) {
429 file_oomem(ms, psize);
434 #if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH)
439 size_t bytesconsumed;
441 (void)memset(&state, 0, sizeof(mbstate_t));
448 bytesconsumed = mbrtowc(&nextchar, op,
449 CAST(size_t, eop - op), &state);
450 if (bytesconsumed == CAST(size_t, -1) ||
451 bytesconsumed == CAST(size_t, -2)) {
456 if (iswprint(nextchar)) {
457 (void)memcpy(np, op, bytesconsumed);
461 while (bytesconsumed-- > 0)
467 /* Parsing succeeded as a multi-byte sequence */
473 for (np = ms->o.pbuf, op = ms->o.buf; *op;) {
474 if (isprint(CAST(unsigned char, *op))) {
485 file_check_mem(struct magic_set *ms, unsigned int level)
489 if (level >= ms->c.len) {
490 len = (ms->c.len = 20 + level) * sizeof(*ms->c.li);
491 ms->c.li = CAST(struct level_info *, (ms->c.li == NULL) ?
493 realloc(ms->c.li, len));
494 if (ms->c.li == NULL) {
499 ms->c.li[level].got_match = 0;
500 #ifdef ENABLE_CONDITIONALS
501 ms->c.li[level].last_match = 0;
502 ms->c.li[level].last_cond = COND_NONE;
503 #endif /* ENABLE_CONDITIONALS */
508 file_printedlen(const struct magic_set *ms)
510 return ms->o.buf == NULL ? 0 : strlen(ms->o.buf);
514 file_replace(struct magic_set *ms, const char *pat, const char *rep)
519 rc = file_regcomp(&rx, pat, REG_EXTENDED);
521 file_regerror(&rx, rc, ms);
525 while (file_regexec(&rx, ms->o.buf, 1, &rm, 0) == 0) {
526 ms->o.buf[rm.rm_so] = '\0';
527 if (file_printf(ms, "%s%s", rep,
528 rm.rm_eo != 0 ? ms->o.buf + rm.rm_eo : "") == -1)
540 file_regcomp(file_regex_t *rx, const char *pat, int flags)
543 rx->c_lc_ctype = newlocale(LC_CTYPE_MASK, "C", 0);
544 assert(rx->c_lc_ctype != NULL);
545 rx->old_lc_ctype = uselocale(rx->c_lc_ctype);
546 assert(rx->old_lc_ctype != NULL);
548 rx->old_lc_ctype = setlocale(LC_CTYPE, "C");
552 return rx->rc = regcomp(&rx->rx, pat, flags);
556 file_regexec(file_regex_t *rx, const char *str, size_t nmatch,
557 regmatch_t* pmatch, int eflags)
560 /* XXX: force initialization because glibc does not always do this */
561 memset(pmatch, 0, nmatch * sizeof(*pmatch));
562 return regexec(&rx->rx, str, nmatch, pmatch, eflags);
566 file_regfree(file_regex_t *rx)
571 (void)uselocale(rx->old_lc_ctype);
572 freelocale(rx->c_lc_ctype);
574 (void)setlocale(LC_CTYPE, rx->old_lc_ctype);
579 file_regerror(file_regex_t *rx, int rc, struct magic_set *ms)
583 (void)regerror(rc, &rx->rx, errmsg, sizeof(errmsg));
584 file_magerror(ms, "regex error %d for `%s', (%s)", rc, rx->pat,
588 protected file_pushbuf_t *
589 file_push_buffer(struct magic_set *ms)
593 if (ms->event_flags & EVENT_HAD_ERR)
596 if ((pb = (CAST(file_pushbuf_t *, malloc(sizeof(*pb))))) == NULL)
600 pb->offset = ms->offset;
609 file_pop_buffer(struct magic_set *ms, file_pushbuf_t *pb)
613 if (ms->event_flags & EVENT_HAD_ERR) {
622 ms->offset = pb->offset;
629 * convert string to ascii printable format.
632 file_printable(char *buf, size_t bufsiz, const char *str, size_t slen)
634 char *ptr, *eptr = buf + bufsiz - 1;
635 const unsigned char *s = RCAST(const unsigned char *, str);
636 const unsigned char *es = s + slen;
638 for (ptr = buf; ptr < eptr && s < es && *s; s++) {
646 *ptr++ = ((CAST(unsigned int, *s) >> 6) & 7) + '0';
647 *ptr++ = ((CAST(unsigned int, *s) >> 3) & 7) + '0';
648 *ptr++ = ((CAST(unsigned int, *s) >> 0) & 7) + '0';