1 /* 2 * Copyright (c) Ian F. Darwin 1986-1995. 3 * Software written by Ian F. Darwin and others; 4 * maintained 1995-present by Christos Zoulas and others. 5 * 6 * Redistribution and use in source and binary forms, with or without 7 * modification, are permitted provided that the following conditions 8 * are met: 9 * 1. Redistributions of source code must retain the above copyright 10 * notice immediately at the beginning of the file, without modification, 11 * this list of conditions, and the following disclaimer. 12 * 2. Redistributions in binary form must reproduce the above copyright 13 * notice, this list of conditions and the following disclaimer in the 14 * documentation and/or other materials provided with the distribution. 15 * 16 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND 17 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 18 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 19 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE FOR 20 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 21 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 22 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 23 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 24 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 25 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 26 * SUCH DAMAGE. 27 */ 28 /* 29 * ASCII magic -- try to detect text encoding. 30 * 31 * Extensively modified by Eric Fischer <enf@pobox.com> in July, 2000, 32 * to handle character codes other than ASCII on a unified basis. 33 */ 34 35 #include "file.h" 36 37 #ifndef lint 38 FILE_RCSID("@(#)$File: ascmagic.c,v 1.81 2011/03/15 22:16:29 christos Exp $") 39 #endif /* lint */ 40 41 #include "magic.h" 42 #include <string.h> 43 #include <memory.h> 44 #include <ctype.h> 45 #include <stdlib.h> 46 #ifdef HAVE_UNISTD_H 47 #include <unistd.h> 48 #endif 49 #include "names.h" 50 51 #define MAXLINELEN 300 /* longest sane line length */ 52 #define ISSPC(x) ((x) == ' ' || (x) == '\t' || (x) == '\r' || (x) == '\n' \ 53 || (x) == 0x85 || (x) == '\f') 54 55 private int ascmatch(const unsigned char *, const unichar *, size_t); 56 private unsigned char *encode_utf8(unsigned char *, size_t, unichar *, size_t); 57 private size_t trim_nuls(const unsigned char *, size_t); 58 59 /* 60 * Undo the NUL-termination kindly provided by process() 61 * but leave at least one byte to look at 62 */ 63 private size_t 64 trim_nuls(const unsigned char *buf, size_t nbytes) 65 { 66 while (nbytes > 1 && buf[nbytes - 1] == '\0') 67 nbytes--; 68 69 return nbytes; 70 } 71 72 protected int 73 file_ascmagic(struct magic_set *ms, const unsigned char *buf, size_t nbytes) 74 { 75 unichar *ubuf = NULL; 76 size_t ulen; 77 int rv = 1; 78 79 const char *code = NULL; 80 const char *code_mime = NULL; 81 const char *type = NULL; 82 83 if (ms->flags & MAGIC_APPLE) 84 return 0; 85 86 nbytes = trim_nuls(buf, nbytes); 87 88 /* If file doesn't look like any sort of text, give up. */ 89 if (file_encoding(ms, buf, nbytes, &ubuf, &ulen, &code, &code_mime, 90 &type) == 0) { 91 rv = 0; 92 goto done; 93 } 94 95 rv = file_ascmagic_with_encoding(ms, buf, nbytes, ubuf, ulen, code, 96 type); 97 98 done: 99 if (ubuf) 100 free(ubuf); 101 102 return rv; 103 } 104 105 protected int 106 file_ascmagic_with_encoding(struct magic_set *ms, const unsigned char *buf, 107 size_t nbytes, unichar *ubuf, size_t ulen, const char *code, 108 const char *type) 109 { 110 unsigned char *utf8_buf = NULL, *utf8_end; 111 size_t mlen, i; 112 const struct names *p; 113 int rv = -1; 114 int mime = ms->flags & MAGIC_MIME; 115 116 const char *subtype = NULL; 117 const char *subtype_mime = NULL; 118 119 int has_escapes = 0; 120 int has_backspace = 0; 121 int seen_cr = 0; 122 123 int n_crlf = 0; 124 int n_lf = 0; 125 int n_cr = 0; 126 int n_nel = 0; 127 int score, curtype, executable = 0; 128 129 size_t last_line_end = (size_t)-1; 130 int has_long_lines = 0; 131 132 if (ms->flags & MAGIC_APPLE) 133 return 0; 134 135 nbytes = trim_nuls(buf, nbytes); 136 137 /* If we have fewer than 2 bytes, give up. */ 138 if (nbytes <= 1) { 139 rv = 0; 140 goto done; 141 } 142 143 if ((ms->flags & MAGIC_NO_CHECK_SOFT) == 0) { 144 /* Convert ubuf to UTF-8 and try text soft magic */ 145 /* malloc size is a conservative overestimate; could be 146 improved, or at least realloced after conversion. */ 147 mlen = ulen * 6; 148 if ((utf8_buf = CAST(unsigned char *, malloc(mlen))) == NULL) { 149 file_oomem(ms, mlen); 150 goto done; 151 } 152 if ((utf8_end = encode_utf8(utf8_buf, mlen, ubuf, ulen)) 153 == NULL) 154 goto done; 155 if ((rv = file_softmagic(ms, utf8_buf, 156 (size_t)(utf8_end - utf8_buf), TEXTTEST)) != 0) 157 goto subtype_identified; 158 else 159 rv = -1; 160 } 161 162 /* look for tokens from names.h - this is expensive! */ 163 if ((ms->flags & MAGIC_NO_CHECK_TOKENS) != 0) 164 goto subtype_identified; 165 166 i = 0; 167 score = 0; 168 curtype = -1; 169 while (i < ulen) { 170 size_t end; 171 172 /* skip past any leading space */ 173 while (i < ulen && ISSPC(ubuf[i])) 174 i++; 175 if (i >= ulen) 176 break; 177 178 /* find the next whitespace */ 179 for (end = i + 1; end < nbytes; end++) 180 if (ISSPC(ubuf[end])) 181 break; 182 183 /* compare the word thus isolated against the token list */ 184 for (p = names; p < names + NNAMES; p++) { 185 if (ascmatch((const unsigned char *)p->name, ubuf + i, 186 end - i)) { 187 if (curtype == -1) 188 curtype = p->type; 189 else if (curtype != p->type) { 190 score = p->score; 191 curtype = p->type; 192 } else 193 score += p->score; 194 if (score > 1) { 195 subtype = types[p->type].human; 196 subtype_mime = types[p->type].mime; 197 goto subtype_identified; 198 } 199 } 200 } 201 202 i = end; 203 } 204 205 subtype_identified: 206 207 /* Now try to discover other details about the file. */ 208 for (i = 0; i < ulen; i++) { 209 if (ubuf[i] == '\n') { 210 if (seen_cr) 211 n_crlf++; 212 else 213 n_lf++; 214 last_line_end = i; 215 } else if (seen_cr) 216 n_cr++; 217 218 seen_cr = (ubuf[i] == '\r'); 219 if (seen_cr) 220 last_line_end = i; 221 222 if (ubuf[i] == 0x85) { /* X3.64/ECMA-43 "next line" character */ 223 n_nel++; 224 last_line_end = i; 225 } 226 227 /* If this line is _longer_ than MAXLINELEN, remember it. */ 228 if (i > last_line_end + MAXLINELEN) 229 has_long_lines = 1; 230 231 if (ubuf[i] == '\033') 232 has_escapes = 1; 233 if (ubuf[i] == '\b') 234 has_backspace = 1; 235 } 236 237 /* Beware, if the data has been truncated, the final CR could have 238 been followed by a LF. If we have HOWMANY bytes, it indicates 239 that the data might have been truncated, probably even before 240 this function was called. */ 241 if (seen_cr && nbytes < HOWMANY) 242 n_cr++; 243 244 if (strcmp(type, "binary") == 0) { 245 rv = 0; 246 goto done; 247 } 248 if (mime) { 249 if (!file_printedlen(ms) && (mime & MAGIC_MIME_TYPE) != 0) { 250 if (subtype_mime) { 251 if (file_printf(ms, "%s", subtype_mime) == -1) 252 goto done; 253 } else { 254 if (file_printf(ms, "text/plain") == -1) 255 goto done; 256 } 257 } 258 } else { 259 if (file_printedlen(ms)) { 260 switch (file_replace(ms, " text$", ", ")) { 261 case 0: 262 switch (file_replace(ms, " text executable$", 263 ", ")) { 264 case 0: 265 if (file_printf(ms, ", ") == -1) 266 goto done; 267 case -1: 268 goto done; 269 default: 270 executable = 1; 271 break; 272 } 273 break; 274 case -1: 275 goto done; 276 default: 277 break; 278 } 279 } 280 281 if (file_printf(ms, "%s", code) == -1) 282 goto done; 283 284 if (subtype) { 285 if (file_printf(ms, " %s", subtype) == -1) 286 goto done; 287 } 288 289 if (file_printf(ms, " %s", type) == -1) 290 goto done; 291 292 if (executable) 293 if (file_printf(ms, " executable") == -1) 294 goto done; 295 296 if (has_long_lines) 297 if (file_printf(ms, ", with very long lines") == -1) 298 goto done; 299 300 /* 301 * Only report line terminators if we find one other than LF, 302 * or if we find none at all. 303 */ 304 if ((n_crlf == 0 && n_cr == 0 && n_nel == 0 && n_lf == 0) || 305 (n_crlf != 0 || n_cr != 0 || n_nel != 0)) { 306 if (file_printf(ms, ", with") == -1) 307 goto done; 308 309 if (n_crlf == 0 && n_cr == 0 && n_nel == 0 && n_lf == 0) { 310 if (file_printf(ms, " no") == -1) 311 goto done; 312 } else { 313 if (n_crlf) { 314 if (file_printf(ms, " CRLF") == -1) 315 goto done; 316 if (n_cr || n_lf || n_nel) 317 if (file_printf(ms, ",") == -1) 318 goto done; 319 } 320 if (n_cr) { 321 if (file_printf(ms, " CR") == -1) 322 goto done; 323 if (n_lf || n_nel) 324 if (file_printf(ms, ",") == -1) 325 goto done; 326 } 327 if (n_lf) { 328 if (file_printf(ms, " LF") == -1) 329 goto done; 330 if (n_nel) 331 if (file_printf(ms, ",") == -1) 332 goto done; 333 } 334 if (n_nel) 335 if (file_printf(ms, " NEL") == -1) 336 goto done; 337 } 338 339 if (file_printf(ms, " line terminators") == -1) 340 goto done; 341 } 342 343 if (has_escapes) 344 if (file_printf(ms, ", with escape sequences") == -1) 345 goto done; 346 if (has_backspace) 347 if (file_printf(ms, ", with overstriking") == -1) 348 goto done; 349 } 350 rv = 1; 351 done: 352 if (utf8_buf) 353 free(utf8_buf); 354 355 return rv; 356 } 357 358 private int 359 ascmatch(const unsigned char *s, const unichar *us, size_t ulen) 360 { 361 size_t i; 362 363 for (i = 0; i < ulen; i++) { 364 if (s[i] != us[i]) 365 return 0; 366 } 367 368 if (s[i]) 369 return 0; 370 else 371 return 1; 372 } 373 374 /* 375 * Encode Unicode string as UTF-8, returning pointer to character 376 * after end of string, or NULL if an invalid character is found. 377 */ 378 private unsigned char * 379 encode_utf8(unsigned char *buf, size_t len, unichar *ubuf, size_t ulen) 380 { 381 size_t i; 382 unsigned char *end = buf + len; 383 384 for (i = 0; i < ulen; i++) { 385 if (ubuf[i] <= 0x7f) { 386 if (end - buf < 1) 387 return NULL; 388 *buf++ = (unsigned char)ubuf[i]; 389 } else if (ubuf[i] <= 0x7ff) { 390 if (end - buf < 2) 391 return NULL; 392 *buf++ = (unsigned char)((ubuf[i] >> 6) + 0xc0); 393 *buf++ = (unsigned char)((ubuf[i] & 0x3f) + 0x80); 394 } else if (ubuf[i] <= 0xffff) { 395 if (end - buf < 3) 396 return NULL; 397 *buf++ = (unsigned char)((ubuf[i] >> 12) + 0xe0); 398 *buf++ = (unsigned char)(((ubuf[i] >> 6) & 0x3f) + 0x80); 399 *buf++ = (unsigned char)((ubuf[i] & 0x3f) + 0x80); 400 } else if (ubuf[i] <= 0x1fffff) { 401 if (end - buf < 4) 402 return NULL; 403 *buf++ = (unsigned char)((ubuf[i] >> 18) + 0xf0); 404 *buf++ = (unsigned char)(((ubuf[i] >> 12) & 0x3f) + 0x80); 405 *buf++ = (unsigned char)(((ubuf[i] >> 6) & 0x3f) + 0x80); 406 *buf++ = (unsigned char)((ubuf[i] & 0x3f) + 0x80); 407 } else if (ubuf[i] <= 0x3ffffff) { 408 if (end - buf < 5) 409 return NULL; 410 *buf++ = (unsigned char)((ubuf[i] >> 24) + 0xf8); 411 *buf++ = (unsigned char)(((ubuf[i] >> 18) & 0x3f) + 0x80); 412 *buf++ = (unsigned char)(((ubuf[i] >> 12) & 0x3f) + 0x80); 413 *buf++ = (unsigned char)(((ubuf[i] >> 6) & 0x3f) + 0x80); 414 *buf++ = (unsigned char)((ubuf[i] & 0x3f) + 0x80); 415 } else if (ubuf[i] <= 0x7fffffff) { 416 if (end - buf < 6) 417 return NULL; 418 *buf++ = (unsigned char)((ubuf[i] >> 30) + 0xfc); 419 *buf++ = (unsigned char)(((ubuf[i] >> 24) & 0x3f) + 0x80); 420 *buf++ = (unsigned char)(((ubuf[i] >> 18) & 0x3f) + 0x80); 421 *buf++ = (unsigned char)(((ubuf[i] >> 12) & 0x3f) + 0x80); 422 *buf++ = (unsigned char)(((ubuf[i] >> 6) & 0x3f) + 0x80); 423 *buf++ = (unsigned char)((ubuf[i] & 0x3f) + 0x80); 424 } else /* Invalid character */ 425 return NULL; 426 } 427 428 return buf; 429 } 430