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 1.84 2015/09/10 13:32:19 christos Exp $")
39 #if defined(HAVE_WCHAR_H)
42 #if defined(HAVE_WCTYPE_H)
45 #if defined(HAVE_LIMITS_H)
50 #define SIZE_MAX ((size_t)~0)
54 * Like printf, only we append to a buffer.
57 file_vprintf(struct magic_set *ms, const char *fmt, va_list ap)
62 if (ms->event_flags & EVENT_HAD_ERR)
64 len = vasprintf(&buf, fmt, ap);
68 if (ms->o.buf != NULL) {
69 len = asprintf(&newstr, "%s%s", ms->o.buf, buf);
79 file_error(ms, errno, "vasprintf failed");
84 file_printf(struct magic_set *ms, const char *fmt, ...)
90 rv = file_vprintf(ms, fmt, ap);
96 * error - print best error message possible
99 __attribute__((__format__(__printf__, 3, 0)))
101 file_error_core(struct magic_set *ms, int error, const char *f, va_list va,
104 /* Only the first error is ok */
105 if (ms->event_flags & EVENT_HAD_ERR)
110 file_printf(ms, "line %" SIZE_T_FORMAT "u:", lineno);
112 if (ms->o.buf && *ms->o.buf)
113 file_printf(ms, " ");
114 file_vprintf(ms, f, va);
116 file_printf(ms, " (%s)", strerror(error));
117 ms->event_flags |= EVENT_HAD_ERR;
123 file_error(struct magic_set *ms, int error, const char *f, ...)
127 file_error_core(ms, error, f, va, 0);
132 * Print an error with magic line number.
136 file_magerror(struct magic_set *ms, const char *f, ...)
140 file_error_core(ms, 0, f, va, ms->line);
145 file_oomem(struct magic_set *ms, size_t len)
147 file_error(ms, errno, "cannot allocate %" SIZE_T_FORMAT "u bytes",
152 file_badseek(struct magic_set *ms)
154 file_error(ms, errno, "error seeking");
158 file_badread(struct magic_set *ms)
160 file_error(ms, errno, "error reading");
166 checkdone(struct magic_set *ms, int *rv)
168 if ((ms->flags & MAGIC_CONTINUE) == 0)
170 if (file_printf(ms, "\n- ") == -1)
177 file_buffer(struct magic_set *ms, int fd, const char *inname __attribute__ ((__unused__)),
178 const void *buf, size_t nb)
180 int m = 0, rv = 0, looks_text = 0;
181 int mime = ms->flags & MAGIC_MIME;
182 const unsigned char *ubuf = CAST(const unsigned char *, buf);
183 unichar *u8buf = NULL;
185 const char *code = NULL;
186 const char *code_mime = "binary";
187 const char *type = "application/octet-stream";
188 const char *def = "data";
189 const char *ftype = NULL;
193 type = "application/x-empty";
195 } else if (nb == 1) {
196 def = "very short file (no magic)";
200 if ((ms->flags & MAGIC_NO_CHECK_ENCODING) == 0) {
201 looks_text = file_encoding(ms, ubuf, nb, &u8buf, &ulen,
202 &code, &code_mime, &ftype);
206 if ((ms->flags & MAGIC_NO_CHECK_APPTYPE) == 0 && inname) {
207 m = file_os2_apptype(ms, inname, buf, nb);
208 if ((ms->flags & MAGIC_DEBUG) != 0)
209 (void)fprintf(stderr, "[try os2_apptype %d]\n", m);
221 /* try compression stuff */
222 if ((ms->flags & MAGIC_NO_CHECK_COMPRESS) == 0) {
223 m = file_zmagic(ms, fd, inname, ubuf, nb);
224 if ((ms->flags & MAGIC_DEBUG) != 0)
225 (void)fprintf(stderr, "[try zmagic %d]\n", m);
231 /* Check if we have a tar file */
232 if ((ms->flags & MAGIC_NO_CHECK_TAR) == 0) {
233 m = file_is_tar(ms, ubuf, nb);
234 if ((ms->flags & MAGIC_DEBUG) != 0)
235 (void)fprintf(stderr, "[try tar %d]\n", m);
237 if (checkdone(ms, &rv))
242 /* Check if we have a CDF file */
243 if ((ms->flags & MAGIC_NO_CHECK_CDF) == 0) {
244 m = file_trycdf(ms, fd, ubuf, nb);
245 if ((ms->flags & MAGIC_DEBUG) != 0)
246 (void)fprintf(stderr, "[try cdf %d]\n", m);
248 if (checkdone(ms, &rv))
253 /* try soft magic tests */
254 if ((ms->flags & MAGIC_NO_CHECK_SOFT) == 0)
255 m = file_softmagic(ms, ubuf, nb, 0, NULL, BINTEST, looks_text);
256 if ((ms->flags & MAGIC_DEBUG) != 0)
257 (void)fprintf(stderr, "[try softmagic %d]\n", m);
260 if ((ms->flags & MAGIC_NO_CHECK_ELF) == 0 && m == 1 &&
261 nb > 5 && fd != -1) {
263 * We matched something in the file, so this
264 * *might* be an ELF file, and the file is at
265 * least 5 bytes long, so if it's an ELF file
266 * it has at least one byte past the ELF magic
267 * number - try extracting information from the
268 * ELF headers that cannot easily * be
269 * extracted with rules in the magic file.
271 m = file_tryelf(ms, fd, ubuf, nb);
272 if ((ms->flags & MAGIC_DEBUG) != 0)
273 (void)fprintf(stderr, "[try elf %d]\n",
277 if (checkdone(ms, &rv))
281 /* try text properties */
282 if ((ms->flags & MAGIC_NO_CHECK_TEXT) == 0) {
284 m = file_ascmagic(ms, ubuf, nb, looks_text);
285 if ((ms->flags & MAGIC_DEBUG) != 0)
286 (void)fprintf(stderr, "[try ascmagic %d]\n", m);
288 if (checkdone(ms, &rv))
296 if ((!mime || (mime & MAGIC_MIME_TYPE)) &&
297 file_printf(ms, "%s", mime ? type : def) == -1) {
301 if ((ms->flags & MAGIC_MIME_ENCODING) != 0) {
302 if (ms->flags & MAGIC_MIME_TYPE)
303 if (file_printf(ms, "; charset=") == -1)
305 if (file_printf(ms, "%s", code_mime) == -1)
320 file_reset(struct magic_set *ms)
322 if (ms->mlist[0] == NULL) {
323 file_error(ms, 0, "no magic files loaded");
334 ms->event_flags &= ~EVENT_HAD_ERR;
339 #define OCTALIFY(n, o) \
341 (void)(*(n)++ = '\\', \
342 *(n)++ = (((uint32_t)*(o) >> 6) & 3) + '0', \
343 *(n)++ = (((uint32_t)*(o) >> 3) & 7) + '0', \
344 *(n)++ = (((uint32_t)*(o) >> 0) & 7) + '0', \
347 protected const char *
348 file_getbuffer(struct magic_set *ms)
350 char *pbuf, *op, *np;
353 if (ms->event_flags & EVENT_HAD_ERR)
356 if (ms->flags & MAGIC_RAW)
359 if (ms->o.buf == NULL)
362 /* * 4 is for octal representation, + 1 is for NUL */
363 len = strlen(ms->o.buf);
364 if (len > (SIZE_MAX - 1) / 4) {
369 if ((pbuf = CAST(char *, realloc(ms->o.pbuf, psize))) == NULL) {
370 file_oomem(ms, psize);
375 #if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH)
380 size_t bytesconsumed;
382 (void)memset(&state, 0, sizeof(mbstate_t));
389 bytesconsumed = mbrtowc(&nextchar, op,
390 (size_t)(eop - op), &state);
391 if (bytesconsumed == (size_t)(-1) ||
392 bytesconsumed == (size_t)(-2)) {
397 if (iswprint(nextchar)) {
398 (void)memcpy(np, op, bytesconsumed);
402 while (bytesconsumed-- > 0)
408 /* Parsing succeeded as a multi-byte sequence */
414 for (np = ms->o.pbuf, op = ms->o.buf; *op;) {
415 if (isprint((unsigned char)*op)) {
426 file_check_mem(struct magic_set *ms, unsigned int level)
430 if (level >= ms->c.len) {
431 len = (ms->c.len = 20 + level) * sizeof(*ms->c.li);
432 ms->c.li = CAST(struct level_info *, (ms->c.li == NULL) ?
434 realloc(ms->c.li, len));
435 if (ms->c.li == NULL) {
440 ms->c.li[level].got_match = 0;
441 #ifdef ENABLE_CONDITIONALS
442 ms->c.li[level].last_match = 0;
443 ms->c.li[level].last_cond = COND_NONE;
444 #endif /* ENABLE_CONDITIONALS */
449 file_printedlen(const struct magic_set *ms)
451 return ms->o.buf == NULL ? 0 : strlen(ms->o.buf);
455 file_replace(struct magic_set *ms, const char *pat, const char *rep)
460 rc = file_regcomp(&rx, pat, REG_EXTENDED);
462 file_regerror(&rx, rc, ms);
466 while (file_regexec(&rx, ms->o.buf, 1, &rm, 0) == 0) {
467 ms->o.buf[rm.rm_so] = '\0';
468 if (file_printf(ms, "%s%s", rep,
469 rm.rm_eo != 0 ? ms->o.buf + rm.rm_eo : "") == -1)
481 file_regcomp(file_regex_t *rx, const char *pat, int flags)
484 rx->c_lc_ctype = newlocale(LC_CTYPE_MASK, "C", 0);
485 assert(rx->c_lc_ctype != NULL);
486 rx->old_lc_ctype = uselocale(rx->c_lc_ctype);
487 assert(rx->old_lc_ctype != NULL);
491 return rx->rc = regcomp(&rx->rx, pat, flags);
495 file_regexec(file_regex_t *rx, const char *str, size_t nmatch,
496 regmatch_t* pmatch, int eflags)
499 return regexec(&rx->rx, str, nmatch, pmatch, eflags);
503 file_regfree(file_regex_t *rx)
508 (void)uselocale(rx->old_lc_ctype);
509 freelocale(rx->c_lc_ctype);
514 file_regerror(file_regex_t *rx, int rc, struct magic_set *ms)
518 (void)regerror(rc, &rx->rx, errmsg, sizeof(errmsg));
519 file_magerror(ms, "regex error %d for `%s', (%s)", rc, rx->pat,
523 protected file_pushbuf_t *
524 file_push_buffer(struct magic_set *ms)
528 if (ms->event_flags & EVENT_HAD_ERR)
531 if ((pb = (CAST(file_pushbuf_t *, malloc(sizeof(*pb))))) == NULL)
535 pb->offset = ms->offset;
544 file_pop_buffer(struct magic_set *ms, file_pushbuf_t *pb)
548 if (ms->event_flags & EVENT_HAD_ERR) {
557 ms->offset = pb->offset;
564 * convert string to ascii printable format.
567 file_printable(char *buf, size_t bufsiz, const char *str)
570 const unsigned char *s = (const unsigned char *)str;
572 for (ptr = buf, eptr = ptr + bufsiz - 1; ptr < eptr && *s; s++) {
580 *ptr++ = ((CAST(unsigned int, *s) >> 6) & 7) + '0';
581 *ptr++ = ((CAST(unsigned int, *s) >> 3) & 7) + '0';
582 *ptr++ = ((CAST(unsigned int, *s) >> 0) & 7) + '0';