funcs.c 9.8 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426
  1. /*
  2. * Copyright (c) Christos Zoulas 2003.
  3. * All Rights Reserved.
  4. *
  5. * Redistribution and use in source and binary forms, with or without
  6. * modification, are permitted provided that the following conditions
  7. * are met:
  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.
  14. *
  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
  25. * SUCH DAMAGE.
  26. */
  27. #include "file.h"
  28. #ifndef lint
  29. FILE_RCSID("@(#)$File: funcs.c,v 1.53 2009/04/07 11:07:00 christos Exp $")
  30. #endif /* lint */
  31. #include "magic.h"
  32. #include <stdarg.h>
  33. #include <stdlib.h>
  34. #include <string.h>
  35. #include <ctype.h>
  36. #if defined(HAVE_WCHAR_H)
  37. #include <wchar.h>
  38. #endif
  39. #if defined(HAVE_WCTYPE_H)
  40. #include <wctype.h>
  41. #endif
  42. #if defined(HAVE_LIMITS_H)
  43. #include <limits.h>
  44. #endif
  45. #ifndef SIZE_MAX
  46. #define SIZE_MAX ((size_t)~0)
  47. #endif
  48. /*
  49. * Like printf, only we append to a buffer.
  50. */
  51. protected int
  52. file_vprintf(struct magic_set *ms, const char *fmt, va_list ap)
  53. {
  54. int len;
  55. char *buf, *newstr;
  56. len = vasprintf(&buf, fmt, ap);
  57. if (len < 0)
  58. goto out;
  59. if (ms->o.buf != NULL) {
  60. len = asprintf(&newstr, "%s%s", ms->o.buf, buf);
  61. free(buf);
  62. if (len < 0)
  63. goto out;
  64. free(ms->o.buf);
  65. buf = newstr;
  66. }
  67. ms->o.buf = buf;
  68. return 0;
  69. out:
  70. file_error(ms, errno, "vasprintf failed");
  71. return -1;
  72. }
  73. protected int
  74. file_printf(struct magic_set *ms, const char *fmt, ...)
  75. {
  76. int rv;
  77. va_list ap;
  78. va_start(ap, fmt);
  79. rv = file_vprintf(ms, fmt, ap);
  80. va_end(ap);
  81. return rv;
  82. }
  83. /*
  84. * error - print best error message possible
  85. */
  86. /*VARARGS*/
  87. private void
  88. file_error_core(struct magic_set *ms, int error, const char *f, va_list va,
  89. uint32_t lineno)
  90. {
  91. /* Only the first error is ok */
  92. if (ms->event_flags & EVENT_HAD_ERR)
  93. return;
  94. if (lineno != 0) {
  95. free(ms->o.buf);
  96. ms->o.buf = NULL;
  97. file_printf(ms, "line %u: ", lineno);
  98. }
  99. file_vprintf(ms, f, va);
  100. if (error > 0)
  101. file_printf(ms, " (%s)", strerror(error));
  102. ms->event_flags |= EVENT_HAD_ERR;
  103. ms->error = error;
  104. }
  105. /*VARARGS*/
  106. protected void
  107. file_error(struct magic_set *ms, int error, const char *f, ...)
  108. {
  109. va_list va;
  110. va_start(va, f);
  111. file_error_core(ms, error, f, va, 0);
  112. va_end(va);
  113. }
  114. /*
  115. * Print an error with magic line number.
  116. */
  117. /*VARARGS*/
  118. protected void
  119. file_magerror(struct magic_set *ms, const char *f, ...)
  120. {
  121. va_list va;
  122. va_start(va, f);
  123. file_error_core(ms, 0, f, va, ms->line);
  124. va_end(va);
  125. }
  126. protected void
  127. file_oomem(struct magic_set *ms, size_t len)
  128. {
  129. file_error(ms, errno, "cannot allocate %zu bytes", len);
  130. }
  131. protected void
  132. file_badseek(struct magic_set *ms)
  133. {
  134. file_error(ms, errno, "error seeking");
  135. }
  136. protected void
  137. file_badread(struct magic_set *ms)
  138. {
  139. file_error(ms, errno, "error reading");
  140. }
  141. #ifndef COMPILE_ONLY
  142. protected int
  143. file_buffer(struct magic_set *ms, int fd, const char *inname, const void *buf,
  144. size_t nb)
  145. {
  146. int m = 0, rv = 0, looks_text = 0;
  147. int mime = ms->flags & MAGIC_MIME;
  148. const unsigned char *ubuf = CAST(const unsigned char *, buf);
  149. unichar *u8buf = NULL;
  150. size_t ulen;
  151. const char *code = NULL;
  152. const char *code_mime = "binary";
  153. const char *type = NULL;
  154. if (nb == 0) {
  155. if ((!mime || (mime & MAGIC_MIME_TYPE)) &&
  156. file_printf(ms, mime ? "application/x-empty" :
  157. "empty") == -1)
  158. return -1;
  159. return 1;
  160. } else if (nb == 1) {
  161. if ((!mime || (mime & MAGIC_MIME_TYPE)) &&
  162. file_printf(ms, mime ? "application/octet-stream" :
  163. "very short file (no magic)") == -1)
  164. return -1;
  165. return 1;
  166. }
  167. if ((ms->flags & MAGIC_NO_CHECK_ENCODING) == 0) {
  168. looks_text = file_encoding(ms, ubuf, nb, &u8buf, &ulen,
  169. &code, &code_mime, &type);
  170. }
  171. #ifdef __EMX__
  172. if ((ms->flags & MAGIC_NO_CHECK_APPTYPE) == 0 && inname) {
  173. switch (file_os2_apptype(ms, inname, buf, nb)) {
  174. case -1:
  175. return -1;
  176. case 0:
  177. break;
  178. default:
  179. return 1;
  180. }
  181. }
  182. #endif
  183. /* try compression stuff */
  184. if ((ms->flags & MAGIC_NO_CHECK_COMPRESS) == 0)
  185. if ((m = file_zmagic(ms, fd, inname, ubuf, nb)) != 0) {
  186. if ((ms->flags & MAGIC_DEBUG) != 0)
  187. (void)fprintf(stderr, "zmagic %d\n", m);
  188. goto done;
  189. }
  190. /* Check if we have a tar file */
  191. if ((ms->flags & MAGIC_NO_CHECK_TAR) == 0)
  192. if ((m = file_is_tar(ms, ubuf, nb)) != 0) {
  193. if ((ms->flags & MAGIC_DEBUG) != 0)
  194. (void)fprintf(stderr, "tar %d\n", m);
  195. goto done;
  196. }
  197. /* Check if we have a CDF file */
  198. if ((ms->flags & MAGIC_NO_CHECK_CDF) == 0)
  199. if ((m = file_trycdf(ms, fd, ubuf, nb)) != 0) {
  200. if ((ms->flags & MAGIC_DEBUG) != 0)
  201. (void)fprintf(stderr, "cdf %d\n", m);
  202. goto done;
  203. }
  204. /* try soft magic tests */
  205. if ((ms->flags & MAGIC_NO_CHECK_SOFT) == 0)
  206. if ((m = file_softmagic(ms, ubuf, nb, BINTEST)) != 0) {
  207. if ((ms->flags & MAGIC_DEBUG) != 0)
  208. (void)fprintf(stderr, "softmagic %d\n", m);
  209. #ifdef BUILTIN_ELF
  210. if ((ms->flags & MAGIC_NO_CHECK_ELF) == 0 && m == 1 &&
  211. nb > 5 && fd != -1) {
  212. /*
  213. * We matched something in the file, so this
  214. * *might* be an ELF file, and the file is at
  215. * least 5 bytes long, so if it's an ELF file
  216. * it has at least one byte past the ELF magic
  217. * number - try extracting information from the
  218. * ELF headers that cannot easily * be
  219. * extracted with rules in the magic file.
  220. */
  221. if ((m = file_tryelf(ms, fd, ubuf, nb)) != 0)
  222. if ((ms->flags & MAGIC_DEBUG) != 0)
  223. (void)fprintf(stderr,
  224. "elf %d\n", m);
  225. }
  226. #endif
  227. goto done;
  228. }
  229. /* try text properties (and possibly text tokens) */
  230. if ((ms->flags & MAGIC_NO_CHECK_TEXT) == 0) {
  231. if ((m = file_ascmagic(ms, ubuf, nb)) != 0) {
  232. if ((ms->flags & MAGIC_DEBUG) != 0)
  233. (void)fprintf(stderr, "ascmagic %d\n", m);
  234. goto done;
  235. }
  236. /* try to discover text encoding */
  237. if ((ms->flags & MAGIC_NO_CHECK_ENCODING) == 0) {
  238. if (looks_text == 0)
  239. if ((m = file_ascmagic_with_encoding( ms, ubuf,
  240. nb, u8buf, ulen, code, type)) != 0) {
  241. if ((ms->flags & MAGIC_DEBUG) != 0)
  242. (void)fprintf(stderr,
  243. "ascmagic/enc %d\n", m);
  244. goto done;
  245. }
  246. }
  247. }
  248. /* give up */
  249. m = 1;
  250. if ((!mime || (mime & MAGIC_MIME_TYPE)) &&
  251. file_printf(ms, mime ? "application/octet-stream" : "data") == -1) {
  252. rv = -1;
  253. }
  254. done:
  255. if ((ms->flags & MAGIC_MIME_ENCODING) != 0) {
  256. if (ms->flags & MAGIC_MIME_TYPE)
  257. if (file_printf(ms, "; charset=") == -1)
  258. rv = -1;
  259. if (file_printf(ms, "%s", code_mime) == -1)
  260. rv = -1;
  261. }
  262. if (u8buf)
  263. free(u8buf);
  264. if (rv)
  265. return rv;
  266. return m;
  267. }
  268. #endif
  269. protected int
  270. file_reset(struct magic_set *ms)
  271. {
  272. if (ms->mlist == NULL) {
  273. file_error(ms, 0, "no magic files loaded");
  274. return -1;
  275. }
  276. if (ms->o.buf) {
  277. free(ms->o.buf);
  278. ms->o.buf = NULL;
  279. }
  280. if (ms->o.pbuf) {
  281. free(ms->o.pbuf);
  282. ms->o.pbuf = NULL;
  283. }
  284. ms->event_flags &= ~EVENT_HAD_ERR;
  285. ms->error = -1;
  286. return 0;
  287. }
  288. #define OCTALIFY(n, o) \
  289. /*LINTED*/ \
  290. (void)(*(n)++ = '\\', \
  291. *(n)++ = (((uint32_t)*(o) >> 6) & 3) + '0', \
  292. *(n)++ = (((uint32_t)*(o) >> 3) & 7) + '0', \
  293. *(n)++ = (((uint32_t)*(o) >> 0) & 7) + '0', \
  294. (o)++)
  295. protected const char *
  296. file_getbuffer(struct magic_set *ms)
  297. {
  298. char *pbuf, *op, *np;
  299. size_t psize, len;
  300. if (ms->event_flags & EVENT_HAD_ERR)
  301. return NULL;
  302. if (ms->flags & MAGIC_RAW)
  303. return ms->o.buf;
  304. if (ms->o.buf == NULL)
  305. return NULL;
  306. /* * 4 is for octal representation, + 1 is for NUL */
  307. len = strlen(ms->o.buf);
  308. if (len > (SIZE_MAX - 1) / 4) {
  309. file_oomem(ms, len);
  310. return NULL;
  311. }
  312. psize = len * 4 + 1;
  313. if ((pbuf = CAST(char *, realloc(ms->o.pbuf, psize))) == NULL) {
  314. file_oomem(ms, psize);
  315. return NULL;
  316. }
  317. ms->o.pbuf = pbuf;
  318. #if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH)
  319. {
  320. mbstate_t state;
  321. wchar_t nextchar;
  322. int mb_conv = 1;
  323. size_t bytesconsumed;
  324. char *eop;
  325. (void)memset(&state, 0, sizeof(mbstate_t));
  326. np = ms->o.pbuf;
  327. op = ms->o.buf;
  328. eop = op + len;
  329. while (op < eop) {
  330. bytesconsumed = mbrtowc(&nextchar, op,
  331. (size_t)(eop - op), &state);
  332. if (bytesconsumed == (size_t)(-1) ||
  333. bytesconsumed == (size_t)(-2)) {
  334. mb_conv = 0;
  335. break;
  336. }
  337. if (iswprint(nextchar)) {
  338. (void)memcpy(np, op, bytesconsumed);
  339. op += bytesconsumed;
  340. np += bytesconsumed;
  341. } else {
  342. while (bytesconsumed-- > 0)
  343. OCTALIFY(np, op);
  344. }
  345. }
  346. *np = '\0';
  347. /* Parsing succeeded as a multi-byte sequence */
  348. if (mb_conv != 0)
  349. return ms->o.pbuf;
  350. }
  351. #endif
  352. for (np = ms->o.pbuf, op = ms->o.buf; *op; op++) {
  353. if (isprint((unsigned char)*op)) {
  354. *np++ = *op;
  355. } else {
  356. OCTALIFY(np, op);
  357. }
  358. }
  359. *np = '\0';
  360. return ms->o.pbuf;
  361. }
  362. protected int
  363. file_check_mem(struct magic_set *ms, unsigned int level)
  364. {
  365. size_t len;
  366. if (level >= ms->c.len) {
  367. len = (ms->c.len += 20) * sizeof(*ms->c.li);
  368. ms->c.li = CAST(struct level_info *, (ms->c.li == NULL) ?
  369. malloc(len) :
  370. realloc(ms->c.li, len));
  371. if (ms->c.li == NULL) {
  372. file_oomem(ms, len);
  373. return -1;
  374. }
  375. }
  376. ms->c.li[level].got_match = 0;
  377. #ifdef ENABLE_CONDITIONALS
  378. ms->c.li[level].last_match = 0;
  379. ms->c.li[level].last_cond = COND_NONE;
  380. #endif /* ENABLE_CONDITIONALS */
  381. return 0;
  382. }