funcs.c 9.0 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391
  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. #include "magic.h"
  29. #include <stdarg.h>
  30. #include <stdlib.h>
  31. #include <string.h>
  32. #include <ctype.h>
  33. #if defined(HAVE_WCHAR_H)
  34. #include <wchar.h>
  35. #endif
  36. #if defined(HAVE_WCTYPE_H)
  37. #include <wctype.h>
  38. #endif
  39. #if defined(HAVE_LIMITS_H)
  40. #include <limits.h>
  41. #endif
  42. #ifndef SIZE_T_MAX
  43. #ifdef __LP64__
  44. #define SIZE_T_MAX (size_t)0xffffffffffffffffU
  45. #else
  46. #define SIZE_T_MAX (size_t)0xffffffffU
  47. #endif
  48. #endif
  49. #ifndef lint
  50. FILE_RCSID("@(#)$File: funcs.c,v 1.35 2007/12/27 16:35:59 christos Exp $")
  51. #endif /* lint */
  52. #ifndef HAVE_VSNPRINTF
  53. int vsnprintf(char *, size_t, const char *, va_list);
  54. #endif
  55. /*
  56. * Like printf, only we print to a buffer and advance it.
  57. */
  58. protected int
  59. file_printf(struct magic_set *ms, const char *fmt, ...)
  60. {
  61. va_list ap;
  62. size_t size;
  63. ssize_t len;
  64. char *buf;
  65. va_start(ap, fmt);
  66. len = vsnprintf(ms->o.ptr, ms->o.left, fmt, ap);
  67. if (len == -1)
  68. goto out;
  69. if (len >= (ssize_t)ms->o.left) {
  70. long diff; /* XXX: really ptrdiff_t */
  71. va_end(ap);
  72. size = (ms->o.size - ms->o.left) + len + 1024;
  73. if ((buf = realloc(ms->o.buf, size)) == NULL) {
  74. file_oomem(ms, size);
  75. return -1;
  76. }
  77. diff = ms->o.ptr - ms->o.buf;
  78. ms->o.ptr = buf + diff;
  79. ms->o.buf = buf;
  80. ms->o.left = size - diff;
  81. ms->o.size = size;
  82. va_start(ap, fmt);
  83. len = vsnprintf(ms->o.ptr, ms->o.left, fmt, ap);
  84. if (len == -1)
  85. goto out;
  86. }
  87. va_end(ap);
  88. ms->o.ptr += len;
  89. ms->o.left -= len;
  90. return 0;
  91. out:
  92. file_error(ms, errno, "vsnprintf failed");
  93. return -1;
  94. }
  95. /*
  96. * error - print best error message possible
  97. */
  98. /*VARARGS*/
  99. private void
  100. file_error_core(struct magic_set *ms, int error, const char *f, va_list va,
  101. uint32_t lineno)
  102. {
  103. size_t len;
  104. /* Only the first error is ok */
  105. if (ms->haderr)
  106. return;
  107. len = 0;
  108. if (lineno != 0) {
  109. (void)snprintf(ms->o.buf, ms->o.size, "line %u: ", lineno);
  110. len = strlen(ms->o.buf);
  111. }
  112. (void)vsnprintf(ms->o.buf + len, ms->o.size - len, f, va);
  113. if (error > 0) {
  114. len = strlen(ms->o.buf);
  115. (void)snprintf(ms->o.buf + len, ms->o.size - len, " (%s)",
  116. strerror(error));
  117. }
  118. ms->haderr++;
  119. ms->error = error;
  120. }
  121. /*VARARGS*/
  122. protected void
  123. file_error(struct magic_set *ms, int error, const char *f, ...)
  124. {
  125. va_list va;
  126. va_start(va, f);
  127. file_error_core(ms, error, f, va, 0);
  128. va_end(va);
  129. }
  130. /*
  131. * Print an error with magic line number.
  132. */
  133. /*VARARGS*/
  134. protected void
  135. file_magerror(struct magic_set *ms, const char *f, ...)
  136. {
  137. va_list va;
  138. va_start(va, f);
  139. file_error_core(ms, 0, f, va, ms->line);
  140. va_end(va);
  141. }
  142. protected void
  143. file_oomem(struct magic_set *ms, size_t len)
  144. {
  145. file_error(ms, errno, "cannot allocate %zu bytes", len);
  146. }
  147. protected void
  148. file_badseek(struct magic_set *ms)
  149. {
  150. file_error(ms, errno, "error seeking");
  151. }
  152. protected void
  153. file_badread(struct magic_set *ms)
  154. {
  155. file_error(ms, errno, "error reading");
  156. }
  157. #ifndef COMPILE_ONLY
  158. protected int
  159. file_buffer(struct magic_set *ms, int fd, const char *inname, const void *buf,
  160. size_t nb)
  161. {
  162. int m;
  163. int mime = ms->flags & MAGIC_MIME;
  164. if (nb == 0) {
  165. if ((!mime || (mime & MAGIC_MIME_TYPE)) &&
  166. file_printf(ms, mime ? "application/x-empty" :
  167. "empty") == -1)
  168. return -1;
  169. return 1;
  170. } else if (nb == 1) {
  171. if ((!mime || (mime & MAGIC_MIME_TYPE)) &&
  172. file_printf(ms, mime ? "application/octet-stream" :
  173. "very short file (no magic)") == -1)
  174. return -1;
  175. return 1;
  176. }
  177. #ifdef __EMX__
  178. if ((ms->flags & MAGIC_NO_CHECK_APPTYPE) == 0 && inname) {
  179. switch (file_os2_apptype(ms, inname, buf, nb)) {
  180. case -1:
  181. return -1;
  182. case 0:
  183. break;
  184. default:
  185. return 1;
  186. }
  187. }
  188. #endif
  189. /* try compression stuff */
  190. if ((ms->flags & MAGIC_NO_CHECK_COMPRESS) != 0 ||
  191. (m = file_zmagic(ms, fd, inname, buf, nb)) == 0) {
  192. /* Check if we have a tar file */
  193. if ((ms->flags & MAGIC_NO_CHECK_TAR) != 0 ||
  194. (m = file_is_tar(ms, buf, nb)) == 0) {
  195. /* try tests in /etc/magic (or surrogate magic file) */
  196. if ((ms->flags & MAGIC_NO_CHECK_SOFT) != 0 ||
  197. (m = file_softmagic(ms, buf, nb)) == 0) {
  198. /* try known keywords, check whether it is ASCII */
  199. if ((ms->flags & MAGIC_NO_CHECK_ASCII) != 0 ||
  200. (m = file_ascmagic(ms, buf, nb)) == 0) {
  201. /* abandon hope, all ye who remain here */
  202. if ((!mime || (mime & MAGIC_MIME_TYPE)) &&
  203. file_printf(ms, mime ? "application/octet-stream" :
  204. "data") == -1)
  205. return -1;
  206. m = 1;
  207. }
  208. }
  209. }
  210. }
  211. #ifdef BUILTIN_ELF
  212. if ((ms->flags & MAGIC_NO_CHECK_ELF) == 0 && m == 1 &&
  213. nb > 5 && fd != -1) {
  214. /*
  215. * We matched something in the file, so this *might*
  216. * be an ELF file, and the file is at least 5 bytes
  217. * long, so if it's an ELF file it has at least one
  218. * byte past the ELF magic number - try extracting
  219. * information from the ELF headers that cannot easily
  220. * be extracted with rules in the magic file.
  221. */
  222. (void)file_tryelf(ms, fd, buf, nb);
  223. }
  224. #endif
  225. return m;
  226. }
  227. #endif
  228. protected int
  229. file_reset(struct magic_set *ms)
  230. {
  231. if (ms->mlist == NULL) {
  232. file_error(ms, 0, "no magic files loaded");
  233. return -1;
  234. }
  235. ms->o.ptr = ms->o.buf;
  236. ms->o.left = ms->o.size;
  237. ms->haderr = 0;
  238. ms->error = -1;
  239. return 0;
  240. }
  241. #define OCTALIFY(n, o) \
  242. /*LINTED*/ \
  243. (void)(*(n)++ = '\\', \
  244. *(n)++ = (((uint32_t)*(o) >> 6) & 3) + '0', \
  245. *(n)++ = (((uint32_t)*(o) >> 3) & 7) + '0', \
  246. *(n)++ = (((uint32_t)*(o) >> 0) & 7) + '0', \
  247. (o)++)
  248. protected const char *
  249. file_getbuffer(struct magic_set *ms)
  250. {
  251. char *pbuf, *op, *np;
  252. size_t psize, len;
  253. if (ms->haderr)
  254. return NULL;
  255. if (ms->flags & MAGIC_RAW)
  256. return ms->o.buf;
  257. len = ms->o.size - ms->o.left;
  258. /* * 4 is for octal representation, + 1 is for NUL */
  259. if (len > (SIZE_T_MAX - 1) / 4) {
  260. file_oomem(ms, len);
  261. return NULL;
  262. }
  263. psize = len * 4 + 1;
  264. if (ms->o.psize < psize) {
  265. if ((pbuf = realloc(ms->o.pbuf, psize)) == NULL) {
  266. file_oomem(ms, psize);
  267. return NULL;
  268. }
  269. ms->o.psize = psize;
  270. ms->o.pbuf = pbuf;
  271. }
  272. #if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH)
  273. {
  274. mbstate_t state;
  275. wchar_t nextchar;
  276. int mb_conv = 1;
  277. size_t bytesconsumed;
  278. char *eop;
  279. (void)memset(&state, 0, sizeof(mbstate_t));
  280. np = ms->o.pbuf;
  281. op = ms->o.buf;
  282. eop = op + strlen(ms->o.buf);
  283. while (op < eop) {
  284. bytesconsumed = mbrtowc(&nextchar, op,
  285. (size_t)(eop - op), &state);
  286. if (bytesconsumed == (size_t)(-1) ||
  287. bytesconsumed == (size_t)(-2)) {
  288. mb_conv = 0;
  289. break;
  290. }
  291. if (iswprint(nextchar)) {
  292. (void)memcpy(np, op, bytesconsumed);
  293. op += bytesconsumed;
  294. np += bytesconsumed;
  295. } else {
  296. while (bytesconsumed-- > 0)
  297. OCTALIFY(np, op);
  298. }
  299. }
  300. *np = '\0';
  301. /* Parsing succeeded as a multi-byte sequence */
  302. if (mb_conv != 0)
  303. return ms->o.pbuf;
  304. }
  305. #endif
  306. for (np = ms->o.pbuf, op = ms->o.buf; *op; op++) {
  307. if (isprint((unsigned char)*op)) {
  308. *np++ = *op;
  309. } else {
  310. OCTALIFY(np, op);
  311. }
  312. }
  313. *np = '\0';
  314. return ms->o.pbuf;
  315. }
  316. protected int
  317. file_check_mem(struct magic_set *ms, unsigned int level)
  318. {
  319. size_t len;
  320. if (level >= ms->c.len) {
  321. len = (ms->c.len += 20) * sizeof(*ms->c.li);
  322. ms->c.li = (ms->c.li == NULL) ? malloc(len) :
  323. realloc(ms->c.li, len);
  324. if (ms->c.li == NULL) {
  325. file_oomem(ms, len);
  326. return -1;
  327. }
  328. }
  329. ms->c.li[level].got_match = 0;
  330. #ifdef ENABLE_CONDITIONALS
  331. ms->c.li[level].last_match = 0;
  332. ms->c.li[level].last_cond = COND_NONE;
  333. #endif /* ENABLE_CONDITIONALS */
  334. return 0;
  335. }
  336. /*
  337. * Yes these wrappers suffer from buffer overflows, but if your OS does not
  338. * have the real functions, maybe you should consider replacing your OS?
  339. */
  340. #ifndef HAVE_VSNPRINTF
  341. int
  342. vsnprintf(char *buf, size_t len, const char *fmt, va_list ap)
  343. {
  344. return vsprintf(buf, fmt, ap);
  345. }
  346. #endif
  347. #ifndef HAVE_SNPRINTF
  348. /*ARGSUSED*/
  349. int
  350. snprintf(char *buf, size_t len, const char *fmt, ...)
  351. {
  352. int rv;
  353. va_list ap;
  354. va_start(ap, fmt);
  355. rv = vsprintf(buf, fmt, ap);
  356. va_end(ap);
  357. return rv;
  358. }
  359. #endif