1 /*- 2 * Copyright (C) 2009 Gabor Kovesdan <gabor@FreeBSD.org> 3 * Copyright (C) 2012 Oleg Moskalenko <oleg.moskalenko@citrix.com> 4 * All rights reserved. 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, 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 19 * FOR 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 28 #include <sys/cdefs.h> 29 __FBSDID("$FreeBSD$"); 30 31 #include <ctype.h> 32 #include <errno.h> 33 #include <err.h> 34 #include <langinfo.h> 35 #include <math.h> 36 #include <stdlib.h> 37 #include <string.h> 38 #include <wchar.h> 39 #include <wctype.h> 40 41 #include "bwstring.h" 42 #include "sort.h" 43 44 bool byte_sort; 45 46 static wchar_t **wmonths; 47 static unsigned char **cmonths; 48 49 /* initialise months */ 50 51 void 52 initialise_months(void) 53 { 54 const nl_item item[12] = { ABMON_1, ABMON_2, ABMON_3, ABMON_4, 55 ABMON_5, ABMON_6, ABMON_7, ABMON_8, ABMON_9, ABMON_10, 56 ABMON_11, ABMON_12 }; 57 unsigned char *tmp; 58 size_t len; 59 60 if (MB_CUR_MAX == 1) { 61 if (cmonths == NULL) { 62 unsigned char *m; 63 64 cmonths = sort_malloc(sizeof(unsigned char*) * 12); 65 for (int i = 0; i < 12; i++) { 66 cmonths[i] = NULL; 67 tmp = (unsigned char *) nl_langinfo(item[i]); 68 if (tmp == NULL) 69 continue; 70 if (debug_sort) 71 printf("month[%d]=%s\n", i, tmp); 72 len = strlen((char*)tmp); 73 if (len < 1) 74 continue; 75 while (isblank(*tmp)) 76 ++tmp; 77 m = sort_malloc(len + 1); 78 memcpy(m, tmp, len + 1); 79 m[len] = '\0'; 80 for (unsigned int j = 0; j < len; j++) 81 m[j] = toupper(m[j]); 82 cmonths[i] = m; 83 } 84 } 85 86 } else { 87 if (wmonths == NULL) { 88 wchar_t *m; 89 90 wmonths = sort_malloc(sizeof(wchar_t *) * 12); 91 for (int i = 0; i < 12; i++) { 92 wmonths[i] = NULL; 93 tmp = (unsigned char *) nl_langinfo(item[i]); 94 if (tmp == NULL) 95 continue; 96 if (debug_sort) 97 printf("month[%d]=%s\n", i, tmp); 98 len = strlen((char*)tmp); 99 if (len < 1) 100 continue; 101 while (isblank(*tmp)) 102 ++tmp; 103 m = sort_malloc(SIZEOF_WCHAR_STRING(len + 1)); 104 if (mbstowcs(m, (char*)tmp, len) == ((size_t) -1)) 105 continue; 106 m[len] = L'\0'; 107 for (unsigned int j = 0; j < len; j++) 108 m[j] = towupper(m[j]); 109 wmonths[i] = m; 110 } 111 } 112 } 113 } 114 115 /* 116 * Compare two wide-character strings 117 */ 118 static int 119 wide_str_coll(const wchar_t *s1, const wchar_t *s2) 120 { 121 int ret = 0; 122 123 errno = 0; 124 ret = wcscoll(s1, s2); 125 if (errno == EILSEQ) { 126 errno = 0; 127 ret = wcscmp(s1, s2); 128 if (errno != 0) { 129 for (size_t i = 0; ; ++i) { 130 wchar_t c1 = s1[i]; 131 wchar_t c2 = s2[i]; 132 if (c1 == L'\0') 133 return ((c2 == L'\0') ? 0 : -1); 134 if (c2 == L'\0') 135 return (+1); 136 if (c1 == c2) 137 continue; 138 return ((int)(c1 - c2)); 139 } 140 } 141 } 142 return (ret); 143 } 144 145 /* counterparts of wcs functions */ 146 147 void 148 bwsprintf(FILE *f, struct bwstring *bws, const char *prefix, const char *suffix) 149 { 150 151 if (MB_CUR_MAX == 1) 152 fprintf(f, "%s%s%s", prefix, bws->data.cstr, suffix); 153 else 154 fprintf(f, "%s%S%s", prefix, bws->data.wstr, suffix); 155 } 156 157 const void* bwsrawdata(const struct bwstring *bws) 158 { 159 160 return (&(bws->data)); 161 } 162 163 size_t bwsrawlen(const struct bwstring *bws) 164 { 165 166 return ((MB_CUR_MAX == 1) ? bws->len : SIZEOF_WCHAR_STRING(bws->len)); 167 } 168 169 size_t 170 bws_memsize(const struct bwstring *bws) 171 { 172 173 return ((MB_CUR_MAX == 1) ? (bws->len + 2 + sizeof(struct bwstring)) : 174 (SIZEOF_WCHAR_STRING(bws->len + 1) + sizeof(struct bwstring))); 175 } 176 177 void 178 bws_setlen(struct bwstring *bws, size_t newlen) 179 { 180 181 if (bws && newlen != bws->len && newlen <= bws->len) { 182 bws->len = newlen; 183 if (MB_CUR_MAX == 1) 184 bws->data.cstr[newlen] = '\0'; 185 else 186 bws->data.wstr[newlen] = L'\0'; 187 } 188 } 189 190 /* 191 * Allocate a new binary string of specified size 192 */ 193 struct bwstring * 194 bwsalloc(size_t sz) 195 { 196 struct bwstring *ret; 197 198 if (MB_CUR_MAX == 1) 199 ret = sort_malloc(sizeof(struct bwstring) + 1 + sz); 200 else 201 ret = sort_malloc(sizeof(struct bwstring) + 202 SIZEOF_WCHAR_STRING(sz + 1)); 203 ret->len = sz; 204 205 if (MB_CUR_MAX == 1) 206 ret->data.cstr[ret->len] = '\0'; 207 else 208 ret->data.wstr[ret->len] = L'\0'; 209 210 return (ret); 211 } 212 213 /* 214 * Create a copy of binary string. 215 * New string size equals the length of the old string. 216 */ 217 struct bwstring * 218 bwsdup(const struct bwstring *s) 219 { 220 221 if (s == NULL) 222 return (NULL); 223 else { 224 struct bwstring *ret = bwsalloc(s->len); 225 226 if (MB_CUR_MAX == 1) 227 memcpy(ret->data.cstr, s->data.cstr, (s->len)); 228 else 229 memcpy(ret->data.wstr, s->data.wstr, 230 SIZEOF_WCHAR_STRING(s->len)); 231 232 return (ret); 233 } 234 } 235 236 /* 237 * Create a new binary string from a raw binary buffer. 238 */ 239 struct bwstring * 240 bwssbdup(const wchar_t *str, size_t len) 241 { 242 243 if (str == NULL) 244 return ((len == 0) ? bwsalloc(0) : NULL); 245 else { 246 struct bwstring *ret; 247 248 ret = bwsalloc(len); 249 250 if (MB_CUR_MAX == 1) 251 for (size_t i = 0; i < len; ++i) 252 ret->data.cstr[i] = (unsigned char) str[i]; 253 else 254 memcpy(ret->data.wstr, str, SIZEOF_WCHAR_STRING(len)); 255 256 return (ret); 257 } 258 } 259 260 /* 261 * Create a new binary string from a raw binary buffer. 262 */ 263 struct bwstring * 264 bwscsbdup(const unsigned char *str, size_t len) 265 { 266 struct bwstring *ret; 267 268 ret = bwsalloc(len); 269 270 if (str) { 271 if (MB_CUR_MAX == 1) 272 memcpy(ret->data.cstr, str, len); 273 else { 274 mbstate_t mbs; 275 const char *s; 276 size_t charlen, chars, cptr; 277 278 charlen = chars = 0; 279 cptr = 0; 280 s = (const char *) str; 281 282 memset(&mbs, 0, sizeof(mbs)); 283 284 while (cptr < len) { 285 size_t n = MB_CUR_MAX; 286 287 if (n > len - cptr) 288 n = len - cptr; 289 charlen = mbrlen(s + cptr, n, &mbs); 290 switch (charlen) { 291 case 0: 292 /* FALLTHROUGH */ 293 case (size_t) -1: 294 /* FALLTHROUGH */ 295 case (size_t) -2: 296 ret->data.wstr[chars++] = 297 (unsigned char) s[cptr]; 298 ++cptr; 299 break; 300 default: 301 n = mbrtowc(ret->data.wstr + (chars++), 302 s + cptr, charlen, &mbs); 303 if ((n == (size_t)-1) || (n == (size_t)-2)) 304 /* NOTREACHED */ 305 err(2, "mbrtowc error"); 306 cptr += charlen; 307 }; 308 } 309 310 ret->len = chars; 311 ret->data.wstr[ret->len] = L'\0'; 312 } 313 } 314 return (ret); 315 } 316 317 /* 318 * De-allocate object memory 319 */ 320 void 321 bwsfree(const struct bwstring *s) 322 { 323 324 if (s) 325 sort_free(s); 326 } 327 328 /* 329 * Copy content of src binary string to dst. 330 * If the capacity of the dst string is not sufficient, 331 * then the data is truncated. 332 */ 333 size_t 334 bwscpy(struct bwstring *dst, const struct bwstring *src) 335 { 336 size_t nums = src->len; 337 338 if (nums > dst->len) 339 nums = dst->len; 340 dst->len = nums; 341 342 if (MB_CUR_MAX == 1) { 343 memcpy(dst->data.cstr, src->data.cstr, nums); 344 dst->data.cstr[dst->len] = '\0'; 345 } else { 346 memcpy(dst->data.wstr, src->data.wstr, 347 SIZEOF_WCHAR_STRING(nums + 1)); 348 dst->data.wstr[dst->len] = L'\0'; 349 } 350 351 return (nums); 352 } 353 354 /* 355 * Copy content of src binary string to dst, 356 * with specified number of symbols to be copied. 357 * If the capacity of the dst string is not sufficient, 358 * then the data is truncated. 359 */ 360 struct bwstring * 361 bwsncpy(struct bwstring *dst, const struct bwstring *src, size_t size) 362 { 363 size_t nums = src->len; 364 365 if (nums > dst->len) 366 nums = dst->len; 367 if (nums > size) 368 nums = size; 369 dst->len = nums; 370 371 if (MB_CUR_MAX == 1) { 372 memcpy(dst->data.cstr, src->data.cstr, nums); 373 dst->data.cstr[dst->len] = '\0'; 374 } else { 375 memcpy(dst->data.wstr, src->data.wstr, 376 SIZEOF_WCHAR_STRING(nums + 1)); 377 dst->data.wstr[dst->len] = L'\0'; 378 } 379 380 return (dst); 381 } 382 383 /* 384 * Copy content of src binary string to dst, 385 * with specified number of symbols to be copied. 386 * An offset value can be specified, from the start of src string. 387 * If the capacity of the dst string is not sufficient, 388 * then the data is truncated. 389 */ 390 struct bwstring * 391 bwsnocpy(struct bwstring *dst, const struct bwstring *src, size_t offset, 392 size_t size) 393 { 394 395 if (offset >= src->len) { 396 dst->data.wstr[0] = 0; 397 dst->len = 0; 398 } else { 399 size_t nums = src->len - offset; 400 401 if (nums > dst->len) 402 nums = dst->len; 403 if (nums > size) 404 nums = size; 405 dst->len = nums; 406 if (MB_CUR_MAX == 1) { 407 memcpy(dst->data.cstr, src->data.cstr + offset, 408 (nums)); 409 dst->data.cstr[dst->len] = '\0'; 410 } else { 411 memcpy(dst->data.wstr, src->data.wstr + offset, 412 SIZEOF_WCHAR_STRING(nums)); 413 dst->data.wstr[dst->len] = L'\0'; 414 } 415 } 416 return (dst); 417 } 418 419 /* 420 * Write binary string to the file. 421 * The output is ended either with '\n' (nl == true) 422 * or '\0' (nl == false). 423 */ 424 size_t 425 bwsfwrite(struct bwstring *bws, FILE *f, bool zero_ended) 426 { 427 428 if (MB_CUR_MAX == 1) { 429 size_t len = bws->len; 430 431 if (!zero_ended) { 432 bws->data.cstr[len] = '\n'; 433 434 if (fwrite(bws->data.cstr, len + 1, 1, f) < 1) 435 err(2, NULL); 436 437 bws->data.cstr[len] = '\0'; 438 } else if (fwrite(bws->data.cstr, len + 1, 1, f) < 1) 439 err(2, NULL); 440 441 return (len + 1); 442 443 } else { 444 wchar_t eols; 445 size_t printed = 0; 446 447 eols = zero_ended ? btowc('\0') : btowc('\n'); 448 449 while (printed < BWSLEN(bws)) { 450 const wchar_t *s = bws->data.wstr + printed; 451 452 if (*s == L'\0') { 453 int nums; 454 455 nums = fwprintf(f, L"%lc", *s); 456 457 if (nums != 1) 458 err(2, NULL); 459 ++printed; 460 } else { 461 int nums; 462 463 nums = fwprintf(f, L"%ls", s); 464 465 if (nums < 1) 466 err(2, NULL); 467 printed += nums; 468 } 469 } 470 fwprintf(f, L"%lc", eols); 471 return (printed + 1); 472 } 473 } 474 475 /* 476 * Allocate and read a binary string from file. 477 * The strings are nl-ended or zero-ended, depending on the sort setting. 478 */ 479 struct bwstring * 480 bwsfgetln(FILE *f, size_t *len, bool zero_ended, struct reader_buffer *rb) 481 { 482 wint_t eols; 483 484 eols = zero_ended ? btowc('\0') : btowc('\n'); 485 486 if (!zero_ended && (MB_CUR_MAX > 1)) { 487 wchar_t *ret; 488 489 ret = fgetwln(f, len); 490 491 if (ret == NULL) { 492 if (!feof(f)) 493 err(2, NULL); 494 return (NULL); 495 } 496 if (*len > 0) { 497 if (ret[*len - 1] == (wchar_t)eols) 498 --(*len); 499 } 500 return (bwssbdup(ret, *len)); 501 502 } else if (!zero_ended && (MB_CUR_MAX == 1)) { 503 char *ret; 504 505 ret = fgetln(f, len); 506 507 if (ret == NULL) { 508 if (!feof(f)) 509 err(2, NULL); 510 return (NULL); 511 } 512 if (*len > 0) { 513 if (ret[*len - 1] == '\n') 514 --(*len); 515 } 516 return (bwscsbdup((unsigned char*)ret, *len)); 517 518 } else { 519 *len = 0; 520 521 if (feof(f)) 522 return (NULL); 523 524 if (2 >= rb->fgetwln_z_buffer_size) { 525 rb->fgetwln_z_buffer_size += 256; 526 rb->fgetwln_z_buffer = sort_realloc(rb->fgetwln_z_buffer, 527 sizeof(wchar_t) * rb->fgetwln_z_buffer_size); 528 } 529 rb->fgetwln_z_buffer[*len] = 0; 530 531 if (MB_CUR_MAX == 1) 532 while (!feof(f)) { 533 int c; 534 535 c = fgetc(f); 536 537 if (c == EOF) { 538 if (*len == 0) 539 return (NULL); 540 goto line_read_done; 541 } 542 if (c == eols) 543 goto line_read_done; 544 545 if (*len + 1 >= rb->fgetwln_z_buffer_size) { 546 rb->fgetwln_z_buffer_size += 256; 547 rb->fgetwln_z_buffer = sort_realloc(rb->fgetwln_z_buffer, 548 SIZEOF_WCHAR_STRING(rb->fgetwln_z_buffer_size)); 549 } 550 551 rb->fgetwln_z_buffer[*len] = c; 552 rb->fgetwln_z_buffer[++(*len)] = 0; 553 } 554 else 555 while (!feof(f)) { 556 wint_t c = 0; 557 558 c = fgetwc(f); 559 560 if (c == WEOF) { 561 if (*len == 0) 562 return (NULL); 563 goto line_read_done; 564 } 565 if (c == eols) 566 goto line_read_done; 567 568 if (*len + 1 >= rb->fgetwln_z_buffer_size) { 569 rb->fgetwln_z_buffer_size += 256; 570 rb->fgetwln_z_buffer = sort_realloc(rb->fgetwln_z_buffer, 571 SIZEOF_WCHAR_STRING(rb->fgetwln_z_buffer_size)); 572 } 573 574 rb->fgetwln_z_buffer[*len] = c; 575 rb->fgetwln_z_buffer[++(*len)] = 0; 576 } 577 578 line_read_done: 579 /* we do not count the last 0 */ 580 return (bwssbdup(rb->fgetwln_z_buffer, *len)); 581 } 582 } 583 584 int 585 bwsncmp(const struct bwstring *bws1, const struct bwstring *bws2, 586 size_t offset, size_t len) 587 { 588 size_t cmp_len, len1, len2; 589 int res = 0; 590 591 cmp_len = 0; 592 len1 = bws1->len; 593 len2 = bws2->len; 594 595 if (len1 <= offset) { 596 return ((len2 <= offset) ? 0 : -1); 597 } else { 598 if (len2 <= offset) 599 return (+1); 600 else { 601 len1 -= offset; 602 len2 -= offset; 603 604 cmp_len = len1; 605 606 if (len2 < cmp_len) 607 cmp_len = len2; 608 609 if (len < cmp_len) 610 cmp_len = len; 611 612 if (MB_CUR_MAX == 1) { 613 const unsigned char *s1, *s2; 614 615 s1 = bws1->data.cstr + offset; 616 s2 = bws2->data.cstr + offset; 617 618 res = memcmp(s1, s2, cmp_len); 619 620 } else { 621 const wchar_t *s1, *s2; 622 623 s1 = bws1->data.wstr + offset; 624 s2 = bws2->data.wstr + offset; 625 626 res = memcmp(s1, s2, SIZEOF_WCHAR_STRING(cmp_len)); 627 } 628 } 629 } 630 631 if (res == 0) { 632 if (len1 < cmp_len && len1 < len2) 633 res = -1; 634 else if (len2 < cmp_len && len2 < len1) 635 res = +1; 636 } 637 638 return (res); 639 } 640 641 int 642 bwscmp(const struct bwstring *bws1, const struct bwstring *bws2, size_t offset) 643 { 644 size_t len1, len2, cmp_len; 645 int res; 646 647 len1 = bws1->len; 648 len2 = bws2->len; 649 650 len1 -= offset; 651 len2 -= offset; 652 653 cmp_len = len1; 654 655 if (len2 < cmp_len) 656 cmp_len = len2; 657 658 res = bwsncmp(bws1, bws2, offset, cmp_len); 659 660 if (res == 0) { 661 if( len1 < len2) 662 res = -1; 663 else if (len2 < len1) 664 res = +1; 665 } 666 667 return (res); 668 } 669 670 int 671 bws_iterator_cmp(bwstring_iterator iter1, bwstring_iterator iter2, size_t len) 672 { 673 wchar_t c1, c2; 674 size_t i = 0; 675 676 for (i = 0; i < len; ++i) { 677 c1 = bws_get_iter_value(iter1); 678 c2 = bws_get_iter_value(iter2); 679 if (c1 != c2) 680 return (c1 - c2); 681 iter1 = bws_iterator_inc(iter1, 1); 682 iter2 = bws_iterator_inc(iter2, 1); 683 } 684 685 return (0); 686 } 687 688 int 689 bwscoll(const struct bwstring *bws1, const struct bwstring *bws2, size_t offset) 690 { 691 size_t len1, len2; 692 693 len1 = bws1->len; 694 len2 = bws2->len; 695 696 if (len1 <= offset) 697 return ((len2 <= offset) ? 0 : -1); 698 else { 699 if (len2 <= offset) 700 return (+1); 701 else { 702 len1 -= offset; 703 len2 -= offset; 704 705 if (MB_CUR_MAX == 1) { 706 const unsigned char *s1, *s2; 707 708 s1 = bws1->data.cstr + offset; 709 s2 = bws2->data.cstr + offset; 710 711 if (byte_sort) { 712 int res = 0; 713 714 if (len1 > len2) { 715 res = memcmp(s1, s2, len2); 716 if (!res) 717 res = +1; 718 } else if (len1 < len2) { 719 res = memcmp(s1, s2, len1); 720 if (!res) 721 res = -1; 722 } else 723 res = memcmp(s1, s2, len1); 724 725 return (res); 726 727 } else { 728 int res = 0; 729 size_t i, maxlen; 730 731 i = 0; 732 maxlen = len1; 733 734 if (maxlen > len2) 735 maxlen = len2; 736 737 while (i < maxlen) { 738 /* goto next non-zero part: */ 739 while ((i < maxlen) && 740 !s1[i] && !s2[i]) 741 ++i; 742 743 if (i >= maxlen) 744 break; 745 746 if (s1[i] == 0) { 747 if (s2[i] == 0) 748 /* NOTREACHED */ 749 err(2, "bwscoll error 01"); 750 else 751 return (-1); 752 } else if (s2[i] == 0) 753 return (+1); 754 755 res = strcoll((const char*)(s1 + i), (const char*)(s2 + i)); 756 if (res) 757 return (res); 758 759 while ((i < maxlen) && 760 s1[i] && s2[i]) 761 ++i; 762 763 if (i >= maxlen) 764 break; 765 766 if (s1[i] == 0) { 767 if (s2[i] == 0) { 768 ++i; 769 continue; 770 } else 771 return (-1); 772 } else if (s2[i] == 0) 773 return (+1); 774 else 775 /* NOTREACHED */ 776 err(2, "bwscoll error 02"); 777 } 778 779 if (len1 < len2) 780 return (-1); 781 else if (len1 > len2) 782 return (+1); 783 784 return (0); 785 } 786 } else { 787 const wchar_t *s1, *s2; 788 size_t i, maxlen; 789 int res = 0; 790 791 s1 = bws1->data.wstr + offset; 792 s2 = bws2->data.wstr + offset; 793 794 i = 0; 795 maxlen = len1; 796 797 if (maxlen > len2) 798 maxlen = len2; 799 800 while (i < maxlen) { 801 802 /* goto next non-zero part: */ 803 while ((i < maxlen) && 804 !s1[i] && !s2[i]) 805 ++i; 806 807 if (i >= maxlen) 808 break; 809 810 if (s1[i] == 0) { 811 if (s2[i] == 0) 812 /* NOTREACHED */ 813 err(2, "bwscoll error 1"); 814 else 815 return (-1); 816 } else if (s2[i] == 0) 817 return (+1); 818 819 res = wide_str_coll(s1 + i, s2 + i); 820 if (res) 821 return (res); 822 823 while ((i < maxlen) && s1[i] && s2[i]) 824 ++i; 825 826 if (i >= maxlen) 827 break; 828 829 if (s1[i] == 0) { 830 if (s2[i] == 0) { 831 ++i; 832 continue; 833 } else 834 return (-1); 835 } else if (s2[i] == 0) 836 return (+1); 837 else 838 /* NOTREACHED */ 839 err(2, "bwscoll error 2"); 840 } 841 842 if (len1 < len2) 843 return (-1); 844 else if (len1 > len2) 845 return (+1); 846 847 return (0); 848 } 849 } 850 } 851 } 852 853 /* 854 * Correction of the system API 855 */ 856 double 857 bwstod(struct bwstring *s0, bool *empty) 858 { 859 double ret = 0; 860 861 if (MB_CUR_MAX == 1) { 862 unsigned char *end, *s; 863 char *ep; 864 865 s = s0->data.cstr; 866 end = s + s0->len; 867 ep = NULL; 868 869 while (isblank(*s) && s < end) 870 ++s; 871 872 if (!isprint(*s)) { 873 *empty = true; 874 return (0); 875 } 876 877 ret = strtod((char*)s, &ep); 878 if ((unsigned char*) ep == s) { 879 *empty = true; 880 return (0); 881 } 882 } else { 883 wchar_t *end, *ep, *s; 884 885 s = s0->data.wstr; 886 end = s + s0->len; 887 ep = NULL; 888 889 while (iswblank(*s) && s < end) 890 ++s; 891 892 if (!iswprint(*s)) { 893 *empty = true; 894 return (0); 895 } 896 897 ret = wcstod(s, &ep); 898 if (ep == s) { 899 *empty = true; 900 return (0); 901 } 902 } 903 904 *empty = false; 905 return (ret); 906 } 907 908 /* 909 * A helper function for monthcoll. If a line matches 910 * a month name, it returns (number of the month - 1), 911 * while if there is no match, it just return -1. 912 */ 913 914 int 915 bws_month_score(const struct bwstring *s0) 916 { 917 918 if (MB_CUR_MAX == 1) { 919 const unsigned char *end, *s; 920 size_t len; 921 922 s = s0->data.cstr; 923 end = s + s0->len; 924 925 while (isblank(*s) && s < end) 926 ++s; 927 928 len = strlen((const char*)s); 929 930 for (int i = 11; i >= 0; --i) { 931 if (cmonths[i] && 932 (s == (unsigned char*)strstr((const char*)s, (char*)(cmonths[i])))) 933 return (i); 934 } 935 936 } else { 937 const wchar_t *end, *s; 938 size_t len; 939 940 s = s0->data.wstr; 941 end = s + s0->len; 942 943 while (iswblank(*s) && s < end) 944 ++s; 945 946 len = wcslen(s); 947 948 for (int i = 11; i >= 0; --i) { 949 if (wmonths[i] && (s == wcsstr(s, wmonths[i]))) 950 return (i); 951 } 952 } 953 954 return (-1); 955 } 956 957 /* 958 * Rips out leading blanks (-b). 959 */ 960 struct bwstring * 961 ignore_leading_blanks(struct bwstring *str) 962 { 963 964 if (MB_CUR_MAX == 1) { 965 unsigned char *dst, *end, *src; 966 967 src = str->data.cstr; 968 dst = src; 969 end = src + str->len; 970 971 while (src < end && isblank(*src)) 972 ++src; 973 974 if (src != dst) { 975 size_t newlen; 976 977 newlen = BWSLEN(str) - (src - dst); 978 979 while (src < end) { 980 *dst = *src; 981 ++dst; 982 ++src; 983 } 984 bws_setlen(str, newlen); 985 } 986 } else { 987 wchar_t *dst, *end, *src; 988 989 src = str->data.wstr; 990 dst = src; 991 end = src + str->len; 992 993 while (src < end && iswblank(*src)) 994 ++src; 995 996 if (src != dst) { 997 998 size_t newlen = BWSLEN(str) - (src - dst); 999 1000 while (src < end) { 1001 *dst = *src; 1002 ++dst; 1003 ++src; 1004 } 1005 bws_setlen(str, newlen); 1006 1007 } 1008 } 1009 return (str); 1010 } 1011 1012 /* 1013 * Rips out nonprinting characters (-i). 1014 */ 1015 struct bwstring * 1016 ignore_nonprinting(struct bwstring *str) 1017 { 1018 size_t newlen = str->len; 1019 1020 if (MB_CUR_MAX == 1) { 1021 unsigned char *dst, *end, *src; 1022 unsigned char c; 1023 1024 src = str->data.cstr; 1025 dst = src; 1026 end = src + str->len; 1027 1028 while (src < end) { 1029 c = *src; 1030 if (isprint(c)) { 1031 *dst = c; 1032 ++dst; 1033 ++src; 1034 } else { 1035 ++src; 1036 --newlen; 1037 } 1038 } 1039 } else { 1040 wchar_t *dst, *end, *src; 1041 wchar_t c; 1042 1043 src = str->data.wstr; 1044 dst = src; 1045 end = src + str->len; 1046 1047 while (src < end) { 1048 c = *src; 1049 if (iswprint(c)) { 1050 *dst = c; 1051 ++dst; 1052 ++src; 1053 } else { 1054 ++src; 1055 --newlen; 1056 } 1057 } 1058 } 1059 bws_setlen(str, newlen); 1060 1061 return (str); 1062 } 1063 1064 /* 1065 * Rips out any characters that are not alphanumeric characters 1066 * nor blanks (-d). 1067 */ 1068 struct bwstring * 1069 dictionary_order(struct bwstring *str) 1070 { 1071 size_t newlen = str->len; 1072 1073 if (MB_CUR_MAX == 1) { 1074 unsigned char *dst, *end, *src; 1075 unsigned char c; 1076 1077 src = str->data.cstr; 1078 dst = src; 1079 end = src + str->len; 1080 1081 while (src < end) { 1082 c = *src; 1083 if (isalnum(c) || isblank(c)) { 1084 *dst = c; 1085 ++dst; 1086 ++src; 1087 } else { 1088 ++src; 1089 --newlen; 1090 } 1091 } 1092 } else { 1093 wchar_t *dst, *end, *src; 1094 wchar_t c; 1095 1096 src = str->data.wstr; 1097 dst = src; 1098 end = src + str->len; 1099 1100 while (src < end) { 1101 c = *src; 1102 if (iswalnum(c) || iswblank(c)) { 1103 *dst = c; 1104 ++dst; 1105 ++src; 1106 } else { 1107 ++src; 1108 --newlen; 1109 } 1110 } 1111 } 1112 bws_setlen(str, newlen); 1113 1114 return (str); 1115 } 1116 1117 /* 1118 * Converts string to lower case(-f). 1119 */ 1120 struct bwstring * 1121 ignore_case(struct bwstring *str) 1122 { 1123 1124 if (MB_CUR_MAX == 1) { 1125 unsigned char *end, *s; 1126 1127 s = str->data.cstr; 1128 end = s + str->len; 1129 1130 while (s < end) { 1131 *s = toupper(*s); 1132 ++s; 1133 } 1134 } else { 1135 wchar_t *end, *s; 1136 1137 s = str->data.wstr; 1138 end = s + str->len; 1139 1140 while (s < end) { 1141 *s = towupper(*s); 1142 ++s; 1143 } 1144 } 1145 return (str); 1146 } 1147 1148 void 1149 bws_disorder_warnx(struct bwstring *s, const char *fn, size_t pos) 1150 { 1151 1152 if (MB_CUR_MAX == 1) 1153 warnx("%s:%zu: disorder: %s", fn, pos + 1, s->data.cstr); 1154 else 1155 warnx("%s:%zu: disorder: %ls", fn, pos + 1, s->data.wstr); 1156 } 1157