ascmagic.c revision 139368
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 -- file types that we know based on keywords 30 * that can appear anywhere in the file. 31 * 32 * Extensively modified by Eric Fischer <enf@pobox.com> in July, 2000, 33 * to handle character codes other than ASCII on a unified basis. 34 * 35 * Joerg Wunsch <joerg@freebsd.org> wrote the original support for 8-bit 36 * international characters, now subsumed into this file. 37 */ 38 39#include "file.h" 40#include "magic.h" 41#include <stdio.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#ifndef lint 52FILE_RCSID("@(#)$Id: ascmagic.c,v 1.41 2004/09/11 19:15:57 christos Exp $") 53#endif /* lint */ 54 55typedef unsigned long unichar; 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 61private int looks_ascii(const unsigned char *, size_t, unichar *, size_t *); 62private int looks_utf8(const unsigned char *, size_t, unichar *, size_t *); 63private int looks_unicode(const unsigned char *, size_t, unichar *, size_t *); 64private int looks_latin1(const unsigned char *, size_t, unichar *, size_t *); 65private int looks_extended(const unsigned char *, size_t, unichar *, size_t *); 66private void from_ebcdic(const unsigned char *, size_t, unsigned char *); 67private int ascmatch(const unsigned char *, const unichar *, size_t); 68 69 70protected int 71file_ascmagic(struct magic_set *ms, const unsigned char *buf, size_t nbytes) 72{ 73 size_t i; 74 unsigned char nbuf[HOWMANY+1]; /* one extra for terminating '\0' */ 75 unichar ubuf[HOWMANY+1]; /* one extra for terminating '\0' */ 76 size_t ulen; 77 struct names *p; 78 79 const char *code = NULL; 80 const char *code_mime = NULL; 81 const char *type = NULL; 82 const char *subtype = NULL; 83 const char *subtype_mime = NULL; 84 85 int has_escapes = 0; 86 int has_backspace = 0; 87 88 int n_crlf = 0; 89 int n_lf = 0; 90 int n_cr = 0; 91 int n_nel = 0; 92 93 int last_line_end = -1; 94 int has_long_lines = 0; 95 96 /* 97 * Undo the NUL-termination kindly provided by process() 98 * but leave at least one byte to look at 99 */ 100 101 while (nbytes > 1 && buf[nbytes - 1] == '\0') 102 nbytes--; 103 104 /* nbuf and ubuf relies on this */ 105 if (nbytes > HOWMANY) 106 nbytes = HOWMANY; 107 108 /* 109 * Then try to determine whether it's any character code we can 110 * identify. Each of these tests, if it succeeds, will leave 111 * the text converted into one-unichar-per-character Unicode in 112 * ubuf, and the number of characters converted in ulen. 113 */ 114 if (looks_ascii(buf, nbytes, ubuf, &ulen)) { 115 code = "ASCII"; 116 code_mime = "us-ascii"; 117 type = "text"; 118 } else if (looks_utf8(buf, nbytes, ubuf, &ulen)) { 119 code = "UTF-8 Unicode"; 120 code_mime = "utf-8"; 121 type = "text"; 122 } else if ((i = looks_unicode(buf, nbytes, ubuf, &ulen)) != 0) { 123 if (i == 1) 124 code = "Little-endian UTF-16 Unicode"; 125 else 126 code = "Big-endian UTF-16 Unicode"; 127 128 type = "character data"; 129 code_mime = "utf-16"; /* is this defined? */ 130 } else if (looks_latin1(buf, nbytes, ubuf, &ulen)) { 131 code = "ISO-8859"; 132 type = "text"; 133 code_mime = "iso-8859-1"; 134 } else if (looks_extended(buf, nbytes, ubuf, &ulen)) { 135 code = "Non-ISO extended-ASCII"; 136 type = "text"; 137 code_mime = "unknown"; 138 } else { 139 from_ebcdic(buf, nbytes, nbuf); 140 141 if (looks_ascii(nbuf, nbytes, ubuf, &ulen)) { 142 code = "EBCDIC"; 143 type = "character data"; 144 code_mime = "ebcdic"; 145 } else if (looks_latin1(nbuf, nbytes, ubuf, &ulen)) { 146 code = "International EBCDIC"; 147 type = "character data"; 148 code_mime = "ebcdic"; 149 } else { 150 return 0; /* doesn't look like text at all */ 151 } 152 } 153 154 /* 155 * for troff, look for . + letter + letter or .\"; 156 * this must be done to disambiguate tar archives' ./file 157 * and other trash from real troff input. 158 * 159 * I believe Plan 9 troff allows non-ASCII characters in the names 160 * of macros, so this test might possibly fail on such a file. 161 */ 162 if (*ubuf == '.') { 163 unichar *tp = ubuf + 1; 164 165 while (ISSPC(*tp)) 166 ++tp; /* skip leading whitespace */ 167 if ((tp[0] == '\\' && tp[1] == '\"') || 168 (isascii((unsigned char)tp[0]) && 169 isalnum((unsigned char)tp[0]) && 170 isascii((unsigned char)tp[1]) && 171 isalnum((unsigned char)tp[1]) && 172 ISSPC(tp[2]))) { 173 subtype_mime = "text/troff"; 174 subtype = "troff or preprocessor input"; 175 goto subtype_identified; 176 } 177 } 178 179 if ((*buf == 'c' || *buf == 'C') && ISSPC(buf[1])) { 180 subtype_mime = "text/fortran"; 181 subtype = "fortran program"; 182 goto subtype_identified; 183 } 184 185 /* look for tokens from names.h - this is expensive! */ 186 187 i = 0; 188 while (i < ulen) { 189 size_t end; 190 191 /* 192 * skip past any leading space 193 */ 194 while (i < ulen && ISSPC(ubuf[i])) 195 i++; 196 if (i >= ulen) 197 break; 198 199 /* 200 * find the next whitespace 201 */ 202 for (end = i + 1; end < nbytes; end++) 203 if (ISSPC(ubuf[end])) 204 break; 205 206 /* 207 * compare the word thus isolated against the token list 208 */ 209 for (p = names; p < names + NNAMES; p++) { 210 if (ascmatch((const unsigned char *)p->name, ubuf + i, 211 end - i)) { 212 subtype = types[p->type].human; 213 subtype_mime = types[p->type].mime; 214 goto subtype_identified; 215 } 216 } 217 218 i = end; 219 } 220 221subtype_identified: 222 223 /* 224 * Now try to discover other details about the file. 225 */ 226 for (i = 0; i < ulen; i++) { 227 if (i > last_line_end + MAXLINELEN) 228 has_long_lines = 1; 229 230 if (ubuf[i] == '\033') 231 has_escapes = 1; 232 if (ubuf[i] == '\b') 233 has_backspace = 1; 234 235 if (ubuf[i] == '\r' && (i + 1 < ulen && ubuf[i + 1] == '\n')) { 236 n_crlf++; 237 last_line_end = i; 238 } 239 if (ubuf[i] == '\r' && (i + 1 >= ulen || ubuf[i + 1] != '\n')) { 240 n_cr++; 241 last_line_end = i; 242 } 243 if (ubuf[i] == '\n' && ((int)i - 1 < 0 || ubuf[i - 1] != '\r')){ 244 n_lf++; 245 last_line_end = i; 246 } 247 if (ubuf[i] == 0x85) { /* X3.64/ECMA-43 "next line" character */ 248 n_nel++; 249 last_line_end = i; 250 } 251 } 252 253 if ((ms->flags & MAGIC_MIME)) { 254 if (subtype_mime) { 255 if (file_printf(ms, subtype_mime) == -1) 256 return -1; 257 } else { 258 if (file_printf(ms, "text/plain") == -1) 259 return -1; 260 } 261 262 if (code_mime) { 263 if (file_printf(ms, "; charset=") == -1) 264 return -1; 265 if (file_printf(ms, code_mime) == -1) 266 return -1; 267 } 268 } else { 269 if (file_printf(ms, code) == -1) 270 return -1; 271 272 if (subtype) { 273 if (file_printf(ms, " ") == -1) 274 return -1; 275 if (file_printf(ms, subtype) == -1) 276 return -1; 277 } 278 279 if (file_printf(ms, " ") == -1) 280 return -1; 281 if (file_printf(ms, type) == -1) 282 return -1; 283 284 if (has_long_lines) 285 if (file_printf(ms, ", with very long lines") == -1) 286 return -1; 287 288 /* 289 * Only report line terminators if we find one other than LF, 290 * or if we find none at all. 291 */ 292 if ((n_crlf == 0 && n_cr == 0 && n_nel == 0 && n_lf == 0) || 293 (n_crlf != 0 || n_cr != 0 || n_nel != 0)) { 294 if (file_printf(ms, ", with") == -1) 295 return -1; 296 297 if (n_crlf == 0 && n_cr == 0 && n_nel == 0 && n_lf == 0) { 298 if (file_printf(ms, " no") == -1) 299 return -1; 300 } else { 301 if (n_crlf) { 302 if (file_printf(ms, " CRLF") == -1) 303 return -1; 304 if (n_cr || n_lf || n_nel) 305 if (file_printf(ms, ",") == -1) 306 return -1; 307 } 308 if (n_cr) { 309 if (file_printf(ms, " CR") == -1) 310 return -1; 311 if (n_lf || n_nel) 312 if (file_printf(ms, ",") == -1) 313 return -1; 314 } 315 if (n_lf) { 316 if (file_printf(ms, " LF") == -1) 317 return -1; 318 if (n_nel) 319 if (file_printf(ms, ",") == -1) 320 return -1; 321 } 322 if (n_nel) 323 if (file_printf(ms, " NEL") == -1) 324 return -1; 325 } 326 327 if (file_printf(ms, " line terminators") == -1) 328 return -1; 329 } 330 331 if (has_escapes) 332 if (file_printf(ms, ", with escape sequences") == -1) 333 return -1; 334 if (has_backspace) 335 if (file_printf(ms, ", with overstriking") == -1) 336 return -1; 337 } 338 339 return 1; 340} 341 342private int 343ascmatch(const unsigned char *s, const unichar *us, size_t ulen) 344{ 345 size_t i; 346 347 for (i = 0; i < ulen; i++) { 348 if (s[i] != us[i]) 349 return 0; 350 } 351 352 if (s[i]) 353 return 0; 354 else 355 return 1; 356} 357 358/* 359 * This table reflects a particular philosophy about what constitutes 360 * "text," and there is room for disagreement about it. 361 * 362 * Version 3.31 of the file command considered a file to be ASCII if 363 * each of its characters was approved by either the isascii() or 364 * isalpha() function. On most systems, this would mean that any 365 * file consisting only of characters in the range 0x00 ... 0x7F 366 * would be called ASCII text, but many systems might reasonably 367 * consider some characters outside this range to be alphabetic, 368 * so the file command would call such characters ASCII. It might 369 * have been more accurate to call this "considered textual on the 370 * local system" than "ASCII." 371 * 372 * It considered a file to be "International language text" if each 373 * of its characters was either an ASCII printing character (according 374 * to the real ASCII standard, not the above test), a character in 375 * the range 0x80 ... 0xFF, or one of the following control characters: 376 * backspace, tab, line feed, vertical tab, form feed, carriage return, 377 * escape. No attempt was made to determine the language in which files 378 * of this type were written. 379 * 380 * 381 * The table below considers a file to be ASCII if all of its characters 382 * are either ASCII printing characters (again, according to the X3.4 383 * standard, not isascii()) or any of the following controls: bell, 384 * backspace, tab, line feed, form feed, carriage return, esc, nextline. 385 * 386 * I include bell because some programs (particularly shell scripts) 387 * use it literally, even though it is rare in normal text. I exclude 388 * vertical tab because it never seems to be used in real text. I also 389 * include, with hesitation, the X3.64/ECMA-43 control nextline (0x85), 390 * because that's what the dd EBCDIC->ASCII table maps the EBCDIC newline 391 * character to. It might be more appropriate to include it in the 8859 392 * set instead of the ASCII set, but it's got to be included in *something* 393 * we recognize or EBCDIC files aren't going to be considered textual. 394 * Some old Unix source files use SO/SI (^N/^O) to shift between Greek 395 * and Latin characters, so these should possibly be allowed. But they 396 * make a real mess on VT100-style displays if they're not paired properly, 397 * so we are probably better off not calling them text. 398 * 399 * A file is considered to be ISO-8859 text if its characters are all 400 * either ASCII, according to the above definition, or printing characters 401 * from the ISO-8859 8-bit extension, characters 0xA0 ... 0xFF. 402 * 403 * Finally, a file is considered to be international text from some other 404 * character code if its characters are all either ISO-8859 (according to 405 * the above definition) or characters in the range 0x80 ... 0x9F, which 406 * ISO-8859 considers to be control characters but the IBM PC and Macintosh 407 * consider to be printing characters. 408 */ 409 410#define F 0 /* character never appears in text */ 411#define T 1 /* character appears in plain ASCII text */ 412#define I 2 /* character appears in ISO-8859 text */ 413#define X 3 /* character appears in non-ISO extended ASCII (Mac, IBM PC) */ 414 415private char text_chars[256] = { 416 /* BEL BS HT LF FF CR */ 417 F, F, F, F, F, F, F, T, T, T, T, F, T, T, F, F, /* 0x0X */ 418 /* ESC */ 419 F, F, F, F, F, F, F, F, F, F, F, T, F, F, F, F, /* 0x1X */ 420 T, T, T, T, T, T, T, T, T, T, T, T, T, T, T, T, /* 0x2X */ 421 T, T, T, T, T, T, T, T, T, T, T, T, T, T, T, T, /* 0x3X */ 422 T, T, T, T, T, T, T, T, T, T, T, T, T, T, T, T, /* 0x4X */ 423 T, T, T, T, T, T, T, T, T, T, T, T, T, T, T, T, /* 0x5X */ 424 T, T, T, T, T, T, T, T, T, T, T, T, T, T, T, T, /* 0x6X */ 425 T, T, T, T, T, T, T, T, T, T, T, T, T, T, T, F, /* 0x7X */ 426 /* NEL */ 427 X, X, X, X, X, T, X, X, X, X, X, X, X, X, X, X, /* 0x8X */ 428 X, X, X, X, X, X, X, X, X, X, X, X, X, X, X, X, /* 0x9X */ 429 I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, /* 0xaX */ 430 I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, /* 0xbX */ 431 I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, /* 0xcX */ 432 I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, /* 0xdX */ 433 I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, /* 0xeX */ 434 I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I /* 0xfX */ 435}; 436 437private int 438looks_ascii(const unsigned char *buf, size_t nbytes, unichar *ubuf, 439 size_t *ulen) 440{ 441 int i; 442 443 *ulen = 0; 444 445 for (i = 0; i < nbytes; i++) { 446 int t = text_chars[buf[i]]; 447 448 if (t != T) 449 return 0; 450 451 ubuf[(*ulen)++] = buf[i]; 452 } 453 454 return 1; 455} 456 457private int 458looks_latin1(const unsigned char *buf, size_t nbytes, unichar *ubuf, size_t *ulen) 459{ 460 int i; 461 462 *ulen = 0; 463 464 for (i = 0; i < nbytes; i++) { 465 int t = text_chars[buf[i]]; 466 467 if (t != T && t != I) 468 return 0; 469 470 ubuf[(*ulen)++] = buf[i]; 471 } 472 473 return 1; 474} 475 476private int 477looks_extended(const unsigned char *buf, size_t nbytes, unichar *ubuf, 478 size_t *ulen) 479{ 480 int i; 481 482 *ulen = 0; 483 484 for (i = 0; i < nbytes; i++) { 485 int t = text_chars[buf[i]]; 486 487 if (t != T && t != I && t != X) 488 return 0; 489 490 ubuf[(*ulen)++] = buf[i]; 491 } 492 493 return 1; 494} 495 496private int 497looks_utf8(const unsigned char *buf, size_t nbytes, unichar *ubuf, size_t *ulen) 498{ 499 int i, n; 500 unichar c; 501 int gotone = 0; 502 503 *ulen = 0; 504 505 for (i = 0; i < nbytes; i++) { 506 if ((buf[i] & 0x80) == 0) { /* 0xxxxxxx is plain ASCII */ 507 /* 508 * Even if the whole file is valid UTF-8 sequences, 509 * still reject it if it uses weird control characters. 510 */ 511 512 if (text_chars[buf[i]] != T) 513 return 0; 514 515 ubuf[(*ulen)++] = buf[i]; 516 } else if ((buf[i] & 0x40) == 0) { /* 10xxxxxx never 1st byte */ 517 return 0; 518 } else { /* 11xxxxxx begins UTF-8 */ 519 int following; 520 521 if ((buf[i] & 0x20) == 0) { /* 110xxxxx */ 522 c = buf[i] & 0x1f; 523 following = 1; 524 } else if ((buf[i] & 0x10) == 0) { /* 1110xxxx */ 525 c = buf[i] & 0x0f; 526 following = 2; 527 } else if ((buf[i] & 0x08) == 0) { /* 11110xxx */ 528 c = buf[i] & 0x07; 529 following = 3; 530 } else if ((buf[i] & 0x04) == 0) { /* 111110xx */ 531 c = buf[i] & 0x03; 532 following = 4; 533 } else if ((buf[i] & 0x02) == 0) { /* 1111110x */ 534 c = buf[i] & 0x01; 535 following = 5; 536 } else 537 return 0; 538 539 for (n = 0; n < following; n++) { 540 i++; 541 if (i >= nbytes) 542 goto done; 543 544 if ((buf[i] & 0x80) == 0 || (buf[i] & 0x40)) 545 return 0; 546 547 c = (c << 6) + (buf[i] & 0x3f); 548 } 549 550 ubuf[(*ulen)++] = c; 551 gotone = 1; 552 } 553 } 554done: 555 return gotone; /* don't claim it's UTF-8 if it's all 7-bit */ 556} 557 558private int 559looks_unicode(const unsigned char *buf, size_t nbytes, unichar *ubuf, 560 size_t *ulen) 561{ 562 int bigend; 563 int i; 564 565 if (nbytes < 2) 566 return 0; 567 568 if (buf[0] == 0xff && buf[1] == 0xfe) 569 bigend = 0; 570 else if (buf[0] == 0xfe && buf[1] == 0xff) 571 bigend = 1; 572 else 573 return 0; 574 575 *ulen = 0; 576 577 for (i = 2; i + 1 < nbytes; i += 2) { 578 /* XXX fix to properly handle chars > 65536 */ 579 580 if (bigend) 581 ubuf[(*ulen)++] = buf[i + 1] + 256 * buf[i]; 582 else 583 ubuf[(*ulen)++] = buf[i] + 256 * buf[i + 1]; 584 585 if (ubuf[*ulen - 1] == 0xfffe) 586 return 0; 587 if (ubuf[*ulen - 1] < 128 && 588 text_chars[(size_t)ubuf[*ulen - 1]] != T) 589 return 0; 590 } 591 592 return 1 + bigend; 593} 594 595#undef F 596#undef T 597#undef I 598#undef X 599 600/* 601 * This table maps each EBCDIC character to an (8-bit extended) ASCII 602 * character, as specified in the rationale for the dd(1) command in 603 * draft 11.2 (September, 1991) of the POSIX P1003.2 standard. 604 * 605 * Unfortunately it does not seem to correspond exactly to any of the 606 * five variants of EBCDIC documented in IBM's _Enterprise Systems 607 * Architecture/390: Principles of Operation_, SA22-7201-06, Seventh 608 * Edition, July, 1999, pp. I-1 - I-4. 609 * 610 * Fortunately, though, all versions of EBCDIC, including this one, agree 611 * on most of the printing characters that also appear in (7-bit) ASCII. 612 * Of these, only '|', '!', '~', '^', '[', and ']' are in question at all. 613 * 614 * Fortunately too, there is general agreement that codes 0x00 through 615 * 0x3F represent control characters, 0x41 a nonbreaking space, and the 616 * remainder printing characters. 617 * 618 * This is sufficient to allow us to identify EBCDIC text and to distinguish 619 * between old-style and internationalized examples of text. 620 */ 621 622private unsigned char ebcdic_to_ascii[] = { 623 0, 1, 2, 3, 156, 9, 134, 127, 151, 141, 142, 11, 12, 13, 14, 15, 624 16, 17, 18, 19, 157, 133, 8, 135, 24, 25, 146, 143, 28, 29, 30, 31, 625128, 129, 130, 131, 132, 10, 23, 27, 136, 137, 138, 139, 140, 5, 6, 7, 626144, 145, 22, 147, 148, 149, 150, 4, 152, 153, 154, 155, 20, 21, 158, 26, 627' ', 160, 161, 162, 163, 164, 165, 166, 167, 168, 213, '.', '<', '(', '+', '|', 628'&', 169, 170, 171, 172, 173, 174, 175, 176, 177, '!', '$', '*', ')', ';', '~', 629'-', '/', 178, 179, 180, 181, 182, 183, 184, 185, 203, ',', '%', '_', '>', '?', 630186, 187, 188, 189, 190, 191, 192, 193, 194, '`', ':', '#', '@', '\'','=', '"', 631195, 'a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 196, 197, 198, 199, 200, 201, 632202, 'j', 'k', 'l', 'm', 'n', 'o', 'p', 'q', 'r', '^', 204, 205, 206, 207, 208, 633209, 229, 's', 't', 'u', 'v', 'w', 'x', 'y', 'z', 210, 211, 212, '[', 214, 215, 634216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, ']', 230, 231, 635'{', 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 232, 233, 234, 235, 236, 237, 636'}', 'J', 'K', 'L', 'M', 'N', 'O', 'P', 'Q', 'R', 238, 239, 240, 241, 242, 243, 637'\\',159, 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z', 244, 245, 246, 247, 248, 249, 638'0', '1', '2', '3', '4', '5', '6', '7', '8', '9', 250, 251, 252, 253, 254, 255 639}; 640 641#ifdef notdef 642/* 643 * The following EBCDIC-to-ASCII table may relate more closely to reality, 644 * or at least to modern reality. It comes from 645 * 646 * http://ftp.s390.ibm.com/products/oe/bpxqp9.html 647 * 648 * and maps the characters of EBCDIC code page 1047 (the code used for 649 * Unix-derived software on IBM's 390 systems) to the corresponding 650 * characters from ISO 8859-1. 651 * 652 * If this table is used instead of the above one, some of the special 653 * cases for the NEL character can be taken out of the code. 654 */ 655 656private unsigned char ebcdic_1047_to_8859[] = { 6570x00,0x01,0x02,0x03,0x9C,0x09,0x86,0x7F,0x97,0x8D,0x8E,0x0B,0x0C,0x0D,0x0E,0x0F, 6580x10,0x11,0x12,0x13,0x9D,0x0A,0x08,0x87,0x18,0x19,0x92,0x8F,0x1C,0x1D,0x1E,0x1F, 6590x80,0x81,0x82,0x83,0x84,0x85,0x17,0x1B,0x88,0x89,0x8A,0x8B,0x8C,0x05,0x06,0x07, 6600x90,0x91,0x16,0x93,0x94,0x95,0x96,0x04,0x98,0x99,0x9A,0x9B,0x14,0x15,0x9E,0x1A, 6610x20,0xA0,0xE2,0xE4,0xE0,0xE1,0xE3,0xE5,0xE7,0xF1,0xA2,0x2E,0x3C,0x28,0x2B,0x7C, 6620x26,0xE9,0xEA,0xEB,0xE8,0xED,0xEE,0xEF,0xEC,0xDF,0x21,0x24,0x2A,0x29,0x3B,0x5E, 6630x2D,0x2F,0xC2,0xC4,0xC0,0xC1,0xC3,0xC5,0xC7,0xD1,0xA6,0x2C,0x25,0x5F,0x3E,0x3F, 6640xF8,0xC9,0xCA,0xCB,0xC8,0xCD,0xCE,0xCF,0xCC,0x60,0x3A,0x23,0x40,0x27,0x3D,0x22, 6650xD8,0x61,0x62,0x63,0x64,0x65,0x66,0x67,0x68,0x69,0xAB,0xBB,0xF0,0xFD,0xFE,0xB1, 6660xB0,0x6A,0x6B,0x6C,0x6D,0x6E,0x6F,0x70,0x71,0x72,0xAA,0xBA,0xE6,0xB8,0xC6,0xA4, 6670xB5,0x7E,0x73,0x74,0x75,0x76,0x77,0x78,0x79,0x7A,0xA1,0xBF,0xD0,0x5B,0xDE,0xAE, 6680xAC,0xA3,0xA5,0xB7,0xA9,0xA7,0xB6,0xBC,0xBD,0xBE,0xDD,0xA8,0xAF,0x5D,0xB4,0xD7, 6690x7B,0x41,0x42,0x43,0x44,0x45,0x46,0x47,0x48,0x49,0xAD,0xF4,0xF6,0xF2,0xF3,0xF5, 6700x7D,0x4A,0x4B,0x4C,0x4D,0x4E,0x4F,0x50,0x51,0x52,0xB9,0xFB,0xFC,0xF9,0xFA,0xFF, 6710x5C,0xF7,0x53,0x54,0x55,0x56,0x57,0x58,0x59,0x5A,0xB2,0xD4,0xD6,0xD2,0xD3,0xD5, 6720x30,0x31,0x32,0x33,0x34,0x35,0x36,0x37,0x38,0x39,0xB3,0xDB,0xDC,0xD9,0xDA,0x9F 673}; 674#endif 675 676/* 677 * Copy buf[0 ... nbytes-1] into out[], translating EBCDIC to ASCII. 678 */ 679private void 680from_ebcdic(const unsigned char *buf, size_t nbytes, unsigned char *out) 681{ 682 int i; 683 684 for (i = 0; i < nbytes; i++) { 685 out[i] = ebcdic_to_ascii[buf[i]]; 686 } 687} 688