funcs.c 11 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468
  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 1.67 2014/02/12 23:20:53 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. #if defined(HAVE_LOCALE_H)
  46. #include <locale.h>
  47. #endif
  48. #ifndef SIZE_MAX
  49. #define SIZE_MAX ((size_t)~0)
  50. #endif
  51. /*
  52. * Like printf, only we append to a buffer.
  53. */
  54. protected int
  55. file_vprintf(struct magic_set *ms, const char *fmt, va_list ap)
  56. {
  57. int len;
  58. char *buf, *newstr;
  59. if (ms->event_flags & EVENT_HAD_ERR)
  60. return 0;
  61. len = vasprintf(&buf, fmt, ap);
  62. if (len < 0)
  63. goto out;
  64. if (ms->o.buf != NULL) {
  65. len = asprintf(&newstr, "%s%s", ms->o.buf, buf);
  66. free(buf);
  67. if (len < 0)
  68. goto out;
  69. free(ms->o.buf);
  70. buf = newstr;
  71. }
  72. ms->o.buf = buf;
  73. return 0;
  74. out:
  75. file_error(ms, errno, "vasprintf failed");
  76. return -1;
  77. }
  78. protected int
  79. file_printf(struct magic_set *ms, const char *fmt, ...)
  80. {
  81. int rv;
  82. va_list ap;
  83. va_start(ap, fmt);
  84. rv = file_vprintf(ms, fmt, ap);
  85. va_end(ap);
  86. return rv;
  87. }
  88. /*
  89. * error - print best error message possible
  90. */
  91. /*VARARGS*/
  92. __attribute__((__format__(__printf__, 3, 0)))
  93. private void
  94. file_error_core(struct magic_set *ms, int error, const char *f, va_list va,
  95. size_t lineno)
  96. {
  97. /* Only the first error is ok */
  98. if (ms->event_flags & EVENT_HAD_ERR)
  99. return;
  100. if (lineno != 0) {
  101. free(ms->o.buf);
  102. ms->o.buf = NULL;
  103. file_printf(ms, "line %" SIZE_T_FORMAT "u: ", lineno);
  104. }
  105. file_vprintf(ms, f, va);
  106. if (error > 0)
  107. file_printf(ms, " (%s)", strerror(error));
  108. ms->event_flags |= EVENT_HAD_ERR;
  109. ms->error = error;
  110. }
  111. /*VARARGS*/
  112. protected void
  113. file_error(struct magic_set *ms, int error, const char *f, ...)
  114. {
  115. va_list va;
  116. va_start(va, f);
  117. file_error_core(ms, error, f, va, 0);
  118. va_end(va);
  119. }
  120. /*
  121. * Print an error with magic line number.
  122. */
  123. /*VARARGS*/
  124. protected void
  125. file_magerror(struct magic_set *ms, const char *f, ...)
  126. {
  127. va_list va;
  128. va_start(va, f);
  129. file_error_core(ms, 0, f, va, ms->line);
  130. va_end(va);
  131. }
  132. protected void
  133. file_oomem(struct magic_set *ms, size_t len)
  134. {
  135. file_error(ms, errno, "cannot allocate %" SIZE_T_FORMAT "u bytes",
  136. len);
  137. }
  138. protected void
  139. file_badseek(struct magic_set *ms)
  140. {
  141. file_error(ms, errno, "error seeking");
  142. }
  143. protected void
  144. file_badread(struct magic_set *ms)
  145. {
  146. file_error(ms, errno, "error reading");
  147. }
  148. #ifndef COMPILE_ONLY
  149. protected int
  150. file_buffer(struct magic_set *ms, int fd, const char *inname __attribute__ ((unused)),
  151. const void *buf, size_t nb)
  152. {
  153. int m = 0, rv = 0, looks_text = 0;
  154. int mime = ms->flags & MAGIC_MIME;
  155. const unsigned char *ubuf = CAST(const unsigned char *, buf);
  156. unichar *u8buf = NULL;
  157. size_t ulen;
  158. const char *code = NULL;
  159. const char *code_mime = "binary";
  160. const char *type = "application/octet-stream";
  161. const char *def = "data";
  162. if (nb == 0) {
  163. def = "empty";
  164. type = "application/x-empty";
  165. goto simple;
  166. } else if (nb == 1) {
  167. def = "very short file (no magic)";
  168. goto simple;
  169. }
  170. if ((ms->flags & MAGIC_NO_CHECK_ENCODING) == 0) {
  171. looks_text = file_encoding(ms, ubuf, nb, &u8buf, &ulen,
  172. &code, &code_mime, &type);
  173. }
  174. #ifdef __EMX__
  175. if ((ms->flags & MAGIC_NO_CHECK_APPTYPE) == 0 && inname) {
  176. switch (file_os2_apptype(ms, inname, buf, nb)) {
  177. case -1:
  178. return -1;
  179. case 0:
  180. break;
  181. default:
  182. return 1;
  183. }
  184. }
  185. #endif
  186. #if HAVE_FORK
  187. /* try compression stuff */
  188. if ((ms->flags & MAGIC_NO_CHECK_COMPRESS) == 0)
  189. if ((m = file_zmagic(ms, fd, inname, ubuf, nb)) != 0) {
  190. if ((ms->flags & MAGIC_DEBUG) != 0)
  191. (void)fprintf(stderr, "zmagic %d\n", m);
  192. goto done_encoding;
  193. }
  194. #endif
  195. /* Check if we have a tar file */
  196. if ((ms->flags & MAGIC_NO_CHECK_TAR) == 0)
  197. if ((m = file_is_tar(ms, ubuf, nb)) != 0) {
  198. if ((ms->flags & MAGIC_DEBUG) != 0)
  199. (void)fprintf(stderr, "tar %d\n", m);
  200. goto done;
  201. }
  202. /* Check if we have a CDF file */
  203. if ((ms->flags & MAGIC_NO_CHECK_CDF) == 0)
  204. if ((m = file_trycdf(ms, fd, ubuf, nb)) != 0) {
  205. if ((ms->flags & MAGIC_DEBUG) != 0)
  206. (void)fprintf(stderr, "cdf %d\n", m);
  207. goto done;
  208. }
  209. /* try soft magic tests */
  210. if ((ms->flags & MAGIC_NO_CHECK_SOFT) == 0)
  211. if ((m = file_softmagic(ms, ubuf, nb, 0, BINTEST,
  212. looks_text)) != 0) {
  213. if ((ms->flags & MAGIC_DEBUG) != 0)
  214. (void)fprintf(stderr, "softmagic %d\n", m);
  215. #ifdef BUILTIN_ELF
  216. if ((ms->flags & MAGIC_NO_CHECK_ELF) == 0 && m == 1 &&
  217. nb > 5 && fd != -1) {
  218. /*
  219. * We matched something in the file, so this
  220. * *might* be an ELF file, and the file is at
  221. * least 5 bytes long, so if it's an ELF file
  222. * it has at least one byte past the ELF magic
  223. * number - try extracting information from the
  224. * ELF headers that cannot easily * be
  225. * extracted with rules in the magic file.
  226. */
  227. if ((m = file_tryelf(ms, fd, ubuf, nb)) != 0)
  228. if ((ms->flags & MAGIC_DEBUG) != 0)
  229. (void)fprintf(stderr,
  230. "elf %d\n", m);
  231. }
  232. #endif
  233. goto done;
  234. }
  235. /* try text properties */
  236. if ((ms->flags & MAGIC_NO_CHECK_TEXT) == 0) {
  237. if ((m = file_ascmagic(ms, ubuf, nb, looks_text)) != 0) {
  238. if ((ms->flags & MAGIC_DEBUG) != 0)
  239. (void)fprintf(stderr, "ascmagic %d\n", m);
  240. goto done;
  241. }
  242. /* try to discover text encoding */
  243. if ((ms->flags & MAGIC_NO_CHECK_ENCODING) == 0) {
  244. if (looks_text == 0)
  245. if ((m = file_ascmagic_with_encoding( ms, ubuf,
  246. nb, u8buf, ulen, code, type, looks_text))
  247. != 0) {
  248. if ((ms->flags & MAGIC_DEBUG) != 0)
  249. (void)fprintf(stderr,
  250. "ascmagic/enc %d\n", m);
  251. goto done;
  252. }
  253. }
  254. }
  255. simple:
  256. /* give up */
  257. m = 1;
  258. if ((!mime || (mime & MAGIC_MIME_TYPE)) &&
  259. file_printf(ms, "%s", mime ? type : def) == -1) {
  260. rv = -1;
  261. }
  262. done:
  263. if ((ms->flags & MAGIC_MIME_ENCODING) != 0) {
  264. if (ms->flags & MAGIC_MIME_TYPE)
  265. if (file_printf(ms, "; charset=") == -1)
  266. rv = -1;
  267. if (file_printf(ms, "%s", code_mime) == -1)
  268. rv = -1;
  269. }
  270. done_encoding:
  271. free(u8buf);
  272. if (rv)
  273. return rv;
  274. return m;
  275. }
  276. #endif
  277. protected int
  278. file_reset(struct magic_set *ms)
  279. {
  280. if (ms->mlist[0] == NULL) {
  281. file_error(ms, 0, "no magic files loaded");
  282. return -1;
  283. }
  284. if (ms->o.buf) {
  285. free(ms->o.buf);
  286. ms->o.buf = NULL;
  287. }
  288. if (ms->o.pbuf) {
  289. free(ms->o.pbuf);
  290. ms->o.pbuf = NULL;
  291. }
  292. ms->event_flags &= ~EVENT_HAD_ERR;
  293. ms->error = -1;
  294. return 0;
  295. }
  296. #define OCTALIFY(n, o) \
  297. /*LINTED*/ \
  298. (void)(*(n)++ = '\\', \
  299. *(n)++ = (((uint32_t)*(o) >> 6) & 3) + '0', \
  300. *(n)++ = (((uint32_t)*(o) >> 3) & 7) + '0', \
  301. *(n)++ = (((uint32_t)*(o) >> 0) & 7) + '0', \
  302. (o)++)
  303. protected const char *
  304. file_getbuffer(struct magic_set *ms)
  305. {
  306. char *pbuf, *op, *np;
  307. size_t psize, len;
  308. if (ms->event_flags & EVENT_HAD_ERR)
  309. return NULL;
  310. if (ms->flags & MAGIC_RAW)
  311. return ms->o.buf;
  312. if (ms->o.buf == NULL)
  313. return NULL;
  314. /* * 4 is for octal representation, + 1 is for NUL */
  315. len = strlen(ms->o.buf);
  316. if (len > (SIZE_MAX - 1) / 4) {
  317. file_oomem(ms, len);
  318. return NULL;
  319. }
  320. psize = len * 4 + 1;
  321. if ((pbuf = CAST(char *, realloc(ms->o.pbuf, psize))) == NULL) {
  322. file_oomem(ms, psize);
  323. return NULL;
  324. }
  325. ms->o.pbuf = pbuf;
  326. #if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH)
  327. {
  328. mbstate_t state;
  329. wchar_t nextchar;
  330. int mb_conv = 1;
  331. size_t bytesconsumed;
  332. char *eop;
  333. (void)memset(&state, 0, sizeof(mbstate_t));
  334. np = ms->o.pbuf;
  335. op = ms->o.buf;
  336. eop = op + len;
  337. while (op < eop) {
  338. bytesconsumed = mbrtowc(&nextchar, op,
  339. (size_t)(eop - op), &state);
  340. if (bytesconsumed == (size_t)(-1) ||
  341. bytesconsumed == (size_t)(-2)) {
  342. mb_conv = 0;
  343. break;
  344. }
  345. if (iswprint(nextchar)) {
  346. (void)memcpy(np, op, bytesconsumed);
  347. op += bytesconsumed;
  348. np += bytesconsumed;
  349. } else {
  350. while (bytesconsumed-- > 0)
  351. OCTALIFY(np, op);
  352. }
  353. }
  354. *np = '\0';
  355. /* Parsing succeeded as a multi-byte sequence */
  356. if (mb_conv != 0)
  357. return ms->o.pbuf;
  358. }
  359. #endif
  360. for (np = ms->o.pbuf, op = ms->o.buf; *op;) {
  361. if (isprint((unsigned char)*op)) {
  362. *np++ = *op++;
  363. } else {
  364. OCTALIFY(np, op);
  365. }
  366. }
  367. *np = '\0';
  368. return ms->o.pbuf;
  369. }
  370. protected int
  371. file_check_mem(struct magic_set *ms, unsigned int level)
  372. {
  373. size_t len;
  374. if (level >= ms->c.len) {
  375. len = (ms->c.len += 20) * sizeof(*ms->c.li);
  376. ms->c.li = CAST(struct level_info *, (ms->c.li == NULL) ?
  377. malloc(len) :
  378. realloc(ms->c.li, len));
  379. if (ms->c.li == NULL) {
  380. file_oomem(ms, len);
  381. return -1;
  382. }
  383. }
  384. ms->c.li[level].got_match = 0;
  385. #ifdef ENABLE_CONDITIONALS
  386. ms->c.li[level].last_match = 0;
  387. ms->c.li[level].last_cond = COND_NONE;
  388. #endif /* ENABLE_CONDITIONALS */
  389. return 0;
  390. }
  391. protected size_t
  392. file_printedlen(const struct magic_set *ms)
  393. {
  394. return ms->o.buf == NULL ? 0 : strlen(ms->o.buf);
  395. }
  396. protected int
  397. file_replace(struct magic_set *ms, const char *pat, const char *rep)
  398. {
  399. regex_t rx;
  400. int rc, rv = -1;
  401. (void)setlocale(LC_CTYPE, "C");
  402. rc = regcomp(&rx, pat, REG_EXTENDED);
  403. if (rc) {
  404. char errmsg[512];
  405. (void)regerror(rc, &rx, errmsg, sizeof(errmsg));
  406. file_magerror(ms, "regex error %d, (%s)", rc, errmsg);
  407. } else {
  408. regmatch_t rm;
  409. int nm = 0;
  410. while (regexec(&rx, ms->o.buf, 1, &rm, 0) == 0) {
  411. ms->o.buf[rm.rm_so] = '\0';
  412. if (file_printf(ms, "%s%s", rep,
  413. rm.rm_eo != 0 ? ms->o.buf + rm.rm_eo : "") == -1)
  414. goto out;
  415. nm++;
  416. }
  417. regfree(&rx);
  418. rv = nm;
  419. }
  420. out:
  421. (void)setlocale(LC_CTYPE, "");
  422. return rv;
  423. }