1 /* $NetBSD: ascmagic.c,v 1.1.1.2 2011/05/12 20:46:52 christos Exp $ */ 2 3 /* 4 * Copyright (c) Ian F. Darwin 1986-1995. 5 * Software written by Ian F. Darwin and others; 6 * maintained 1995-present by Christos Zoulas and others. 7 * 8 * Redistribution and use in source and binary forms, with or without 9 * modification, are permitted provided that the following conditions 10 * are met: 11 * 1. Redistributions of source code must retain the above copyright 12 * notice immediately at the beginning of the file, without modification, 13 * this list of conditions, and the following disclaimer. 14 * 2. Redistributions in binary form must reproduce the above copyright 15 * notice, this list of conditions and the following disclaimer in the 16 * documentation and/or other materials provided with the distribution. 17 * 18 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND 19 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 20 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 21 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE FOR 22 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 23 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 24 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 25 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 26 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 27 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 28 * SUCH DAMAGE. 29 */ 30 /* 31 * ASCII magic -- try to detect text encoding. 32 * 33 * Extensively modified by Eric Fischer <enf@pobox.com> in July, 2000, 34 * to handle character codes other than ASCII on a unified basis. 35 */ 36 37 #include "file.h" 38 39 #ifndef lint 40 #if 0 41 FILE_RCSID("@(#)$File: ascmagic.c,v 1.81 2011/03/15 22:16:29 christos Exp $") 42 #else 43 __RCSID("$NetBSD: ascmagic.c,v 1.1.1.2 2011/05/12 20:46:52 christos Exp $"); 44 #endif 45 #endif /* lint */ 46 47 #include "magic.h" 48 #include <string.h> 49 #include <memory.h> 50 #include <ctype.h> 51 #include <stdlib.h> 52 #ifdef HAVE_UNISTD_H 53 #include <unistd.h> 54 #endif 55 #include "names.h" 56 57 #define MAXLINELEN 300 /* longest sane line length */ 58 #define ISSPC(x) ((x) == ' ' || (x) == '\t' || (x) == '\r' || (x) == '\n' \ 59 || (x) == 0x85 || (x) == '\f') 60 61 private int ascmatch(const unsigned char *, const unichar *, size_t); 62 private unsigned char *encode_utf8(unsigned char *, size_t, unichar *, size_t); 63 private size_t trim_nuls(const unsigned char *, size_t); 64 65 /* 66 * Undo the NUL-termination kindly provided by process() 67 * but leave at least one byte to look at 68 */ 69 private size_t 70 trim_nuls(const unsigned char *buf, size_t nbytes) 71 { 72 while (nbytes > 1 && buf[nbytes - 1] == '\0') 73 nbytes--; 74 75 return nbytes; 76 } 77 78 protected int 79 file_ascmagic(struct magic_set *ms, const unsigned char *buf, size_t nbytes) 80 { 81 unichar *ubuf = NULL; 82 size_t ulen; 83 int rv = 1; 84 85 const char *code = NULL; 86 const char *code_mime = NULL; 87 const char *type = NULL; 88 89 if (ms->flags & MAGIC_APPLE) 90 return 0; 91 92 nbytes = trim_nuls(buf, nbytes); 93 94 /* If file doesn't look like any sort of text, give up. */ 95 if (file_encoding(ms, buf, nbytes, &ubuf, &ulen, &code, &code_mime, 96 &type) == 0) { 97 rv = 0; 98 goto done; 99 } 100 101 rv = file_ascmagic_with_encoding(ms, buf, nbytes, ubuf, ulen, code, 102 type); 103 104 done: 105 if (ubuf) 106 free(ubuf); 107 108 return rv; 109 } 110 111 protected int 112 file_ascmagic_with_encoding(struct magic_set *ms, const unsigned char *buf, 113 size_t nbytes, unichar *ubuf, size_t ulen, const char *code, 114 const char *type) 115 { 116 unsigned char *utf8_buf = NULL, *utf8_end; 117 size_t mlen, i; 118 const struct names *p; 119 int rv = -1; 120 int mime = ms->flags & MAGIC_MIME; 121 122 const char *subtype = NULL; 123 const char *subtype_mime = NULL; 124 125 int has_escapes = 0; 126 int has_backspace = 0; 127 int seen_cr = 0; 128 129 int n_crlf = 0; 130 int n_lf = 0; 131 int n_cr = 0; 132 int n_nel = 0; 133 int score, curtype, executable = 0; 134 135 size_t last_line_end = (size_t)-1; 136 int has_long_lines = 0; 137 138 if (ms->flags & MAGIC_APPLE) 139 return 0; 140 141 nbytes = trim_nuls(buf, nbytes); 142 143 /* If we have fewer than 2 bytes, give up. */ 144 if (nbytes <= 1) { 145 rv = 0; 146 goto done; 147 } 148 149 if ((ms->flags & MAGIC_NO_CHECK_SOFT) == 0) { 150 /* Convert ubuf to UTF-8 and try text soft magic */ 151 /* malloc size is a conservative overestimate; could be 152 improved, or at least realloced after conversion. */ 153 mlen = ulen * 6; 154 if ((utf8_buf = CAST(unsigned char *, malloc(mlen))) == NULL) { 155 file_oomem(ms, mlen); 156 goto done; 157 } 158 if ((utf8_end = encode_utf8(utf8_buf, mlen, ubuf, ulen)) 159 == NULL) 160 goto done; 161 if ((rv = file_softmagic(ms, utf8_buf, 162 (size_t)(utf8_end - utf8_buf), TEXTTEST)) != 0) 163 goto subtype_identified; 164 else 165 rv = -1; 166 } 167 168 /* look for tokens from names.h - this is expensive! */ 169 if ((ms->flags & MAGIC_NO_CHECK_TOKENS) != 0) 170 goto subtype_identified; 171 172 i = 0; 173 score = 0; 174 curtype = -1; 175 while (i < ulen) { 176 size_t end; 177 178 /* skip past any leading space */ 179 while (i < ulen && ISSPC(ubuf[i])) 180 i++; 181 if (i >= ulen) 182 break; 183 184 /* find the next whitespace */ 185 for (end = i + 1; end < nbytes; end++) 186 if (ISSPC(ubuf[end])) 187 break; 188 189 /* compare the word thus isolated against the token list */ 190 for (p = names; p < names + NNAMES; p++) { 191 if (ascmatch((const unsigned char *)p->name, ubuf + i, 192 end - i)) { 193 if (curtype == -1) 194 curtype = p->type; 195 else if (curtype != p->type) { 196 score = p->score; 197 curtype = p->type; 198 } else 199 score += p->score; 200 if (score > 1) { 201 subtype = types[p->type].human; 202 subtype_mime = types[p->type].mime; 203 goto subtype_identified; 204 } 205 } 206 } 207 208 i = end; 209 } 210 211 subtype_identified: 212 213 /* Now try to discover other details about the file. */ 214 for (i = 0; i < ulen; i++) { 215 if (ubuf[i] == '\n') { 216 if (seen_cr) 217 n_crlf++; 218 else 219 n_lf++; 220 last_line_end = i; 221 } else if (seen_cr) 222 n_cr++; 223 224 seen_cr = (ubuf[i] == '\r'); 225 if (seen_cr) 226 last_line_end = i; 227 228 if (ubuf[i] == 0x85) { /* X3.64/ECMA-43 "next line" character */ 229 n_nel++; 230 last_line_end = i; 231 } 232 233 /* If this line is _longer_ than MAXLINELEN, remember it. */ 234 if (i > last_line_end + MAXLINELEN) 235 has_long_lines = 1; 236 237 if (ubuf[i] == '\033') 238 has_escapes = 1; 239 if (ubuf[i] == '\b') 240 has_backspace = 1; 241 } 242 243 /* Beware, if the data has been truncated, the final CR could have 244 been followed by a LF. If we have HOWMANY bytes, it indicates 245 that the data might have been truncated, probably even before 246 this function was called. */ 247 if (seen_cr && nbytes < HOWMANY) 248 n_cr++; 249 250 if (strcmp(type, "binary") == 0) { 251 rv = 0; 252 goto done; 253 } 254 if (mime) { 255 if (!file_printedlen(ms) && (mime & MAGIC_MIME_TYPE) != 0) { 256 if (subtype_mime) { 257 if (file_printf(ms, "%s", subtype_mime) == -1) 258 goto done; 259 } else { 260 if (file_printf(ms, "text/plain") == -1) 261 goto done; 262 } 263 } 264 } else { 265 if (file_printedlen(ms)) { 266 switch (file_replace(ms, " text$", ", ")) { 267 case 0: 268 switch (file_replace(ms, " text executable$", 269 ", ")) { 270 case 0: 271 if (file_printf(ms, ", ") == -1) 272 goto done; 273 case -1: 274 goto done; 275 default: 276 executable = 1; 277 break; 278 } 279 break; 280 case -1: 281 goto done; 282 default: 283 break; 284 } 285 } 286 287 if (file_printf(ms, "%s", code) == -1) 288 goto done; 289 290 if (subtype) { 291 if (file_printf(ms, " %s", subtype) == -1) 292 goto done; 293 } 294 295 if (file_printf(ms, " %s", type) == -1) 296 goto done; 297 298 if (executable) 299 if (file_printf(ms, " executable") == -1) 300 goto done; 301 302 if (has_long_lines) 303 if (file_printf(ms, ", with very long lines") == -1) 304 goto done; 305 306 /* 307 * Only report line terminators if we find one other than LF, 308 * or if we find none at all. 309 */ 310 if ((n_crlf == 0 && n_cr == 0 && n_nel == 0 && n_lf == 0) || 311 (n_crlf != 0 || n_cr != 0 || n_nel != 0)) { 312 if (file_printf(ms, ", with") == -1) 313 goto done; 314 315 if (n_crlf == 0 && n_cr == 0 && n_nel == 0 && n_lf == 0) { 316 if (file_printf(ms, " no") == -1) 317 goto done; 318 } else { 319 if (n_crlf) { 320 if (file_printf(ms, " CRLF") == -1) 321 goto done; 322 if (n_cr || n_lf || n_nel) 323 if (file_printf(ms, ",") == -1) 324 goto done; 325 } 326 if (n_cr) { 327 if (file_printf(ms, " CR") == -1) 328 goto done; 329 if (n_lf || n_nel) 330 if (file_printf(ms, ",") == -1) 331 goto done; 332 } 333 if (n_lf) { 334 if (file_printf(ms, " LF") == -1) 335 goto done; 336 if (n_nel) 337 if (file_printf(ms, ",") == -1) 338 goto done; 339 } 340 if (n_nel) 341 if (file_printf(ms, " NEL") == -1) 342 goto done; 343 } 344 345 if (file_printf(ms, " line terminators") == -1) 346 goto done; 347 } 348 349 if (has_escapes) 350 if (file_printf(ms, ", with escape sequences") == -1) 351 goto done; 352 if (has_backspace) 353 if (file_printf(ms, ", with overstriking") == -1) 354 goto done; 355 } 356 rv = 1; 357 done: 358 if (utf8_buf) 359 free(utf8_buf); 360 361 return rv; 362 } 363 364 private int 365 ascmatch(const unsigned char *s, const unichar *us, size_t ulen) 366 { 367 size_t i; 368 369 for (i = 0; i < ulen; i++) { 370 if (s[i] != us[i]) 371 return 0; 372 } 373 374 if (s[i]) 375 return 0; 376 else 377 return 1; 378 } 379 380 /* 381 * Encode Unicode string as UTF-8, returning pointer to character 382 * after end of string, or NULL if an invalid character is found. 383 */ 384 private unsigned char * 385 encode_utf8(unsigned char *buf, size_t len, unichar *ubuf, size_t ulen) 386 { 387 size_t i; 388 unsigned char *end = buf + len; 389 390 for (i = 0; i < ulen; i++) { 391 if (ubuf[i] <= 0x7f) { 392 if (end - buf < 1) 393 return NULL; 394 *buf++ = (unsigned char)ubuf[i]; 395 } else if (ubuf[i] <= 0x7ff) { 396 if (end - buf < 2) 397 return NULL; 398 *buf++ = (unsigned char)((ubuf[i] >> 6) + 0xc0); 399 *buf++ = (unsigned char)((ubuf[i] & 0x3f) + 0x80); 400 } else if (ubuf[i] <= 0xffff) { 401 if (end - buf < 3) 402 return NULL; 403 *buf++ = (unsigned char)((ubuf[i] >> 12) + 0xe0); 404 *buf++ = (unsigned char)(((ubuf[i] >> 6) & 0x3f) + 0x80); 405 *buf++ = (unsigned char)((ubuf[i] & 0x3f) + 0x80); 406 } else if (ubuf[i] <= 0x1fffff) { 407 if (end - buf < 4) 408 return NULL; 409 *buf++ = (unsigned char)((ubuf[i] >> 18) + 0xf0); 410 *buf++ = (unsigned char)(((ubuf[i] >> 12) & 0x3f) + 0x80); 411 *buf++ = (unsigned char)(((ubuf[i] >> 6) & 0x3f) + 0x80); 412 *buf++ = (unsigned char)((ubuf[i] & 0x3f) + 0x80); 413 } else if (ubuf[i] <= 0x3ffffff) { 414 if (end - buf < 5) 415 return NULL; 416 *buf++ = (unsigned char)((ubuf[i] >> 24) + 0xf8); 417 *buf++ = (unsigned char)(((ubuf[i] >> 18) & 0x3f) + 0x80); 418 *buf++ = (unsigned char)(((ubuf[i] >> 12) & 0x3f) + 0x80); 419 *buf++ = (unsigned char)(((ubuf[i] >> 6) & 0x3f) + 0x80); 420 *buf++ = (unsigned char)((ubuf[i] & 0x3f) + 0x80); 421 } else if (ubuf[i] <= 0x7fffffff) { 422 if (end - buf < 6) 423 return NULL; 424 *buf++ = (unsigned char)((ubuf[i] >> 30) + 0xfc); 425 *buf++ = (unsigned char)(((ubuf[i] >> 24) & 0x3f) + 0x80); 426 *buf++ = (unsigned char)(((ubuf[i] >> 18) & 0x3f) + 0x80); 427 *buf++ = (unsigned char)(((ubuf[i] >> 12) & 0x3f) + 0x80); 428 *buf++ = (unsigned char)(((ubuf[i] >> 6) & 0x3f) + 0x80); 429 *buf++ = (unsigned char)((ubuf[i] & 0x3f) + 0x80); 430 } else /* Invalid character */ 431 return NULL; 432 } 433 434 return buf; 435 } 436