1 /*- 2 * Copyright (C) 2009 Gabor Kovesdan <gabor@FreeBSD.org> 3 * Copyright (C) 2012 Oleg Moskalenko <mom040267@gmail.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 * $FreeBSD: head/usr.bin/sort/bwstring.c 281181 2015-04-07 01:17:29Z pfg $ 28 */ 29 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 (debug_sort) 69 printf("month[%d]=%s\n", i, tmp); 70 if (*tmp == '\0') 71 continue; 72 m = sort_strdup(tmp); 73 len = strlen(tmp); 74 for (unsigned int j = 0; j < len; j++) 75 m[j] = toupper(m[j]); 76 cmonths[i] = m; 77 } 78 } 79 80 } else { 81 if (wmonths == NULL) { 82 wchar_t *m; 83 84 wmonths = sort_malloc(sizeof(wchar_t *) * 12); 85 for (int i = 0; i < 12; i++) { 86 wmonths[i] = NULL; 87 tmp = (unsigned char *) nl_langinfo(item[i]); 88 if (debug_sort) 89 printf("month[%d]=%s\n", i, tmp); 90 if (*tmp == '\0') 91 continue; 92 len = strlen(tmp); 93 m = sort_malloc(SIZEOF_WCHAR_STRING(len + 1)); 94 if (mbstowcs(m, (char*)tmp, len) == 95 ((size_t) - 1)) { 96 sort_free(m); 97 continue; 98 } 99 m[len] = L'\0'; 100 for (unsigned int j = 0; j < len; j++) 101 m[j] = towupper(m[j]); 102 wmonths[i] = m; 103 } 104 } 105 } 106 } 107 108 /* 109 * Compare two wide-character strings 110 */ 111 static int 112 wide_str_coll(const wchar_t *s1, const wchar_t *s2) 113 { 114 int ret = 0; 115 116 errno = 0; 117 ret = wcscoll(s1, s2); 118 if (errno == EILSEQ) { 119 errno = 0; 120 ret = wcscmp(s1, s2); 121 if (errno != 0) { 122 for (size_t i = 0; ; ++i) { 123 wchar_t c1 = s1[i]; 124 wchar_t c2 = s2[i]; 125 if (c1 == L'\0') 126 return ((c2 == L'\0') ? 0 : -1); 127 if (c2 == L'\0') 128 return (+1); 129 if (c1 == c2) 130 continue; 131 return ((int)(c1 - c2)); 132 } 133 } 134 } 135 return (ret); 136 } 137 138 /* counterparts of wcs functions */ 139 140 void 141 bwsprintf(FILE *f, struct bwstring *bws, const char *prefix, const char *suffix) 142 { 143 144 if (MB_CUR_MAX == 1) 145 fprintf(f, "%s%s%s", prefix, bws->data.cstr, suffix); 146 else 147 fprintf(f, "%s%S%s", prefix, bws->data.wstr, suffix); 148 } 149 150 const void* bwsrawdata(const struct bwstring *bws) 151 { 152 153 return (&(bws->data)); 154 } 155 156 size_t bwsrawlen(const struct bwstring *bws) 157 { 158 159 return ((MB_CUR_MAX == 1) ? bws->len : SIZEOF_WCHAR_STRING(bws->len)); 160 } 161 162 size_t 163 bws_memsize(const struct bwstring *bws) 164 { 165 166 return ((MB_CUR_MAX == 1) ? (bws->len + 2 + sizeof(struct bwstring)) : 167 (SIZEOF_WCHAR_STRING(bws->len + 1) + sizeof(struct bwstring))); 168 } 169 170 void 171 bws_setlen(struct bwstring *bws, size_t newlen) 172 { 173 174 if (bws && newlen != bws->len && newlen <= bws->len) { 175 bws->len = newlen; 176 if (MB_CUR_MAX == 1) 177 bws->data.cstr[newlen] = '\0'; 178 else 179 bws->data.wstr[newlen] = L'\0'; 180 } 181 } 182 183 /* 184 * Allocate a new binary string of specified size 185 */ 186 struct bwstring * 187 bwsalloc(size_t sz) 188 { 189 struct bwstring *ret; 190 191 if (MB_CUR_MAX == 1) 192 ret = sort_malloc(sizeof(struct bwstring) + 1 + sz); 193 else 194 ret = sort_malloc(sizeof(struct bwstring) + 195 SIZEOF_WCHAR_STRING(sz + 1)); 196 ret->len = sz; 197 198 if (MB_CUR_MAX == 1) 199 ret->data.cstr[ret->len] = '\0'; 200 else 201 ret->data.wstr[ret->len] = L'\0'; 202 203 return (ret); 204 } 205 206 /* 207 * Create a copy of binary string. 208 * New string size equals the length of the old string. 209 */ 210 struct bwstring * 211 bwsdup(const struct bwstring *s) 212 { 213 214 if (s == NULL) 215 return (NULL); 216 else { 217 struct bwstring *ret = bwsalloc(s->len); 218 219 if (MB_CUR_MAX == 1) 220 memcpy(ret->data.cstr, s->data.cstr, (s->len)); 221 else 222 memcpy(ret->data.wstr, s->data.wstr, 223 SIZEOF_WCHAR_STRING(s->len)); 224 225 return (ret); 226 } 227 } 228 229 /* 230 * Create a new binary string from a wide character buffer. 231 */ 232 struct bwstring * 233 bwssbdup(const wchar_t *str, size_t len) 234 { 235 236 if (str == NULL) 237 return ((len == 0) ? bwsalloc(0) : NULL); 238 else { 239 struct bwstring *ret; 240 241 ret = bwsalloc(len); 242 243 if (MB_CUR_MAX == 1) 244 for (size_t i = 0; i < len; ++i) 245 ret->data.cstr[i] = (unsigned char) str[i]; 246 else 247 memcpy(ret->data.wstr, str, SIZEOF_WCHAR_STRING(len)); 248 249 return (ret); 250 } 251 } 252 253 /* 254 * Create a new binary string from a raw binary buffer. 255 */ 256 struct bwstring * 257 bwscsbdup(const unsigned char *str, size_t len) 258 { 259 struct bwstring *ret; 260 261 ret = bwsalloc(len); 262 263 if (str) { 264 if (MB_CUR_MAX == 1) 265 memcpy(ret->data.cstr, str, len); 266 else { 267 mbstate_t mbs; 268 const char *s; 269 size_t charlen, chars, cptr; 270 271 charlen = chars = 0; 272 cptr = 0; 273 s = (const char *) str; 274 275 memset(&mbs, 0, sizeof(mbs)); 276 277 while (cptr < len) { 278 size_t n = MB_CUR_MAX; 279 280 if (n > len - cptr) 281 n = len - cptr; 282 charlen = mbrlen(s + cptr, n, &mbs); 283 switch (charlen) { 284 case 0: 285 /* FALLTHROUGH */ 286 case (size_t) -1: 287 /* FALLTHROUGH */ 288 case (size_t) -2: 289 ret->data.wstr[chars++] = 290 (unsigned char) s[cptr]; 291 ++cptr; 292 break; 293 default: 294 n = mbrtowc(ret->data.wstr + (chars++), 295 s + cptr, charlen, &mbs); 296 if ((n == (size_t)-1) || (n == (size_t)-2)) 297 /* NOTREACHED */ 298 err(2, "mbrtowc error"); 299 cptr += charlen; 300 } 301 } 302 303 ret->len = chars; 304 ret->data.wstr[ret->len] = L'\0'; 305 } 306 } 307 return (ret); 308 } 309 310 /* 311 * De-allocate object memory 312 */ 313 void 314 bwsfree(const struct bwstring *s) 315 { 316 317 if (s) 318 sort_free(s); 319 } 320 321 /* 322 * Copy content of src binary string to dst. 323 * If the capacity of the dst string is not sufficient, 324 * then the data is truncated. 325 */ 326 size_t 327 bwscpy(struct bwstring *dst, const struct bwstring *src) 328 { 329 size_t nums = src->len; 330 331 if (nums > dst->len) 332 nums = dst->len; 333 dst->len = nums; 334 335 if (MB_CUR_MAX == 1) { 336 memcpy(dst->data.cstr, src->data.cstr, nums); 337 dst->data.cstr[dst->len] = '\0'; 338 } else { 339 memcpy(dst->data.wstr, src->data.wstr, 340 SIZEOF_WCHAR_STRING(nums + 1)); 341 dst->data.wstr[dst->len] = L'\0'; 342 } 343 344 return (nums); 345 } 346 347 /* 348 * Copy content of src binary string to dst, 349 * with specified number of symbols to be copied. 350 * If the capacity of the dst string is not sufficient, 351 * then the data is truncated. 352 */ 353 struct bwstring * 354 bwsncpy(struct bwstring *dst, const struct bwstring *src, size_t size) 355 { 356 size_t nums = src->len; 357 358 if (nums > dst->len) 359 nums = dst->len; 360 if (nums > size) 361 nums = size; 362 dst->len = nums; 363 364 if (MB_CUR_MAX == 1) { 365 memcpy(dst->data.cstr, src->data.cstr, nums); 366 dst->data.cstr[dst->len] = '\0'; 367 } else { 368 memcpy(dst->data.wstr, src->data.wstr, 369 SIZEOF_WCHAR_STRING(nums + 1)); 370 dst->data.wstr[dst->len] = L'\0'; 371 } 372 373 return (dst); 374 } 375 376 /* 377 * Copy content of src binary string to dst, 378 * with specified number of symbols to be copied. 379 * An offset value can be specified, from the start of src string. 380 * If the capacity of the dst string is not sufficient, 381 * then the data is truncated. 382 */ 383 struct bwstring * 384 bwsnocpy(struct bwstring *dst, const struct bwstring *src, size_t offset, 385 size_t size) 386 { 387 388 if (offset >= src->len) { 389 dst->data.wstr[0] = 0; 390 dst->len = 0; 391 } else { 392 size_t nums = src->len - offset; 393 394 if (nums > dst->len) 395 nums = dst->len; 396 if (nums > size) 397 nums = size; 398 dst->len = nums; 399 if (MB_CUR_MAX == 1) { 400 memcpy(dst->data.cstr, src->data.cstr + offset, 401 (nums)); 402 dst->data.cstr[dst->len] = '\0'; 403 } else { 404 memcpy(dst->data.wstr, src->data.wstr + offset, 405 SIZEOF_WCHAR_STRING(nums)); 406 dst->data.wstr[dst->len] = L'\0'; 407 } 408 } 409 return (dst); 410 } 411 412 /* 413 * Write binary string to the file. 414 * The output is ended either with '\n' (nl == true) 415 * or '\0' (nl == false). 416 */ 417 size_t 418 bwsfwrite(struct bwstring *bws, FILE *f, bool zero_ended) 419 { 420 421 if (MB_CUR_MAX == 1) { 422 size_t len = bws->len; 423 424 if (!zero_ended) { 425 bws->data.cstr[len] = '\n'; 426 427 if (fwrite(bws->data.cstr, len + 1, 1, f) < 1) 428 err(2, NULL); 429 430 bws->data.cstr[len] = '\0'; 431 } else if (fwrite(bws->data.cstr, len + 1, 1, f) < 1) 432 err(2, NULL); 433 434 return (len + 1); 435 436 } else { 437 wchar_t eols; 438 size_t printed = 0; 439 440 eols = zero_ended ? btowc('\0') : btowc('\n'); 441 442 while (printed < BWSLEN(bws)) { 443 const wchar_t *s = bws->data.wstr + printed; 444 445 if (*s == L'\0') { 446 int nums; 447 448 nums = fwprintf(f, L"%lc", *s); 449 450 if (nums != 1) 451 err(2, NULL); 452 ++printed; 453 } else { 454 int nums; 455 456 nums = fwprintf(f, L"%ls", s); 457 458 if (nums < 1) 459 err(2, NULL); 460 printed += nums; 461 } 462 } 463 fwprintf(f, L"%lc", eols); 464 return (printed + 1); 465 } 466 } 467 468 /* 469 * Allocate and read a binary string from file. 470 * The strings are nl-ended or zero-ended, depending on the sort setting. 471 */ 472 struct bwstring * 473 bwsfgetln(FILE *f, size_t *len, bool zero_ended, struct reader_buffer *rb) 474 { 475 wint_t eols; 476 477 eols = zero_ended ? btowc('\0') : btowc('\n'); 478 479 if (!zero_ended && (MB_CUR_MAX > 1)) { 480 wchar_t *ret; 481 482 ret = fgetwln(f, len); 483 484 if (ret == NULL) { 485 if (!feof(f)) 486 err(2, NULL); 487 return (NULL); 488 } 489 if (*len > 0) { 490 if (ret[*len - 1] == (wchar_t)eols) 491 --(*len); 492 } 493 return (bwssbdup(ret, *len)); 494 495 } else if (!zero_ended && (MB_CUR_MAX == 1)) { 496 char *ret; 497 498 ret = fgetln(f, len); 499 500 if (ret == NULL) { 501 if (!feof(f)) 502 err(2, NULL); 503 return (NULL); 504 } 505 if (*len > 0) { 506 if (ret[*len - 1] == '\n') 507 --(*len); 508 } 509 return (bwscsbdup((unsigned char*)ret, *len)); 510 511 } else { 512 *len = 0; 513 514 if (feof(f)) 515 return (NULL); 516 517 if (2 >= rb->fgetwln_z_buffer_size) { 518 rb->fgetwln_z_buffer_size += 256; 519 rb->fgetwln_z_buffer = sort_realloc(rb->fgetwln_z_buffer, 520 sizeof(wchar_t) * rb->fgetwln_z_buffer_size); 521 } 522 rb->fgetwln_z_buffer[*len] = 0; 523 524 if (MB_CUR_MAX == 1) 525 while (!feof(f)) { 526 int c; 527 528 c = fgetc(f); 529 530 if (c == EOF) { 531 if (*len == 0) 532 return (NULL); 533 goto line_read_done; 534 } 535 if (c == eols) 536 goto line_read_done; 537 538 if (*len + 1 >= rb->fgetwln_z_buffer_size) { 539 rb->fgetwln_z_buffer_size += 256; 540 rb->fgetwln_z_buffer = sort_realloc(rb->fgetwln_z_buffer, 541 SIZEOF_WCHAR_STRING(rb->fgetwln_z_buffer_size)); 542 } 543 544 rb->fgetwln_z_buffer[*len] = c; 545 rb->fgetwln_z_buffer[++(*len)] = 0; 546 } 547 else 548 while (!feof(f)) { 549 wint_t c = 0; 550 551 c = fgetwc(f); 552 553 if (c == WEOF) { 554 if (*len == 0) 555 return (NULL); 556 goto line_read_done; 557 } 558 if (c == eols) 559 goto line_read_done; 560 561 if (*len + 1 >= rb->fgetwln_z_buffer_size) { 562 rb->fgetwln_z_buffer_size += 256; 563 rb->fgetwln_z_buffer = sort_realloc(rb->fgetwln_z_buffer, 564 SIZEOF_WCHAR_STRING(rb->fgetwln_z_buffer_size)); 565 } 566 567 rb->fgetwln_z_buffer[*len] = c; 568 rb->fgetwln_z_buffer[++(*len)] = 0; 569 } 570 571 line_read_done: 572 /* we do not count the last 0 */ 573 return (bwssbdup(rb->fgetwln_z_buffer, *len)); 574 } 575 } 576 577 int 578 bwsncmp(const struct bwstring *bws1, const struct bwstring *bws2, 579 size_t offset, size_t len) 580 { 581 size_t cmp_len, len1, len2; 582 int res = 0; 583 584 cmp_len = 0; 585 len1 = bws1->len; 586 len2 = bws2->len; 587 588 if (len1 <= offset) { 589 return ((len2 <= offset) ? 0 : -1); 590 } else { 591 if (len2 <= offset) 592 return (+1); 593 else { 594 len1 -= offset; 595 len2 -= offset; 596 597 cmp_len = len1; 598 599 if (len2 < cmp_len) 600 cmp_len = len2; 601 602 if (len < cmp_len) 603 cmp_len = len; 604 605 if (MB_CUR_MAX == 1) { 606 const unsigned char *s1, *s2; 607 608 s1 = bws1->data.cstr + offset; 609 s2 = bws2->data.cstr + offset; 610 611 res = memcmp(s1, s2, cmp_len); 612 613 } else { 614 const wchar_t *s1, *s2; 615 616 s1 = bws1->data.wstr + offset; 617 s2 = bws2->data.wstr + offset; 618 619 res = memcmp(s1, s2, SIZEOF_WCHAR_STRING(cmp_len)); 620 } 621 } 622 } 623 624 if (res == 0) { 625 if (len1 < cmp_len && len1 < len2) 626 res = -1; 627 else if (len2 < cmp_len && len2 < len1) 628 res = +1; 629 } 630 631 return (res); 632 } 633 634 int 635 bwscmp(const struct bwstring *bws1, const struct bwstring *bws2, size_t offset) 636 { 637 size_t len1, len2, cmp_len; 638 int res; 639 640 len1 = bws1->len; 641 len2 = bws2->len; 642 643 len1 -= offset; 644 len2 -= offset; 645 646 cmp_len = len1; 647 648 if (len2 < cmp_len) 649 cmp_len = len2; 650 651 res = bwsncmp(bws1, bws2, offset, cmp_len); 652 653 if (res == 0) { 654 if( len1 < len2) 655 res = -1; 656 else if (len2 < len1) 657 res = +1; 658 } 659 660 return (res); 661 } 662 663 int 664 bws_iterator_cmp(bwstring_iterator iter1, bwstring_iterator iter2, size_t len) 665 { 666 wchar_t c1, c2; 667 size_t i = 0; 668 669 for (i = 0; i < len; ++i) { 670 c1 = bws_get_iter_value(iter1); 671 c2 = bws_get_iter_value(iter2); 672 if (c1 != c2) 673 return (c1 - c2); 674 iter1 = bws_iterator_inc(iter1, 1); 675 iter2 = bws_iterator_inc(iter2, 1); 676 } 677 678 return (0); 679 } 680 681 int 682 bwscoll(const struct bwstring *bws1, const struct bwstring *bws2, size_t offset) 683 { 684 size_t len1, len2; 685 686 len1 = bws1->len; 687 len2 = bws2->len; 688 689 if (len1 <= offset) 690 return ((len2 <= offset) ? 0 : -1); 691 else { 692 if (len2 <= offset) 693 return (+1); 694 else { 695 len1 -= offset; 696 len2 -= offset; 697 698 if (MB_CUR_MAX == 1) { 699 const unsigned char *s1, *s2; 700 701 s1 = bws1->data.cstr + offset; 702 s2 = bws2->data.cstr + offset; 703 704 if (byte_sort) { 705 int res = 0; 706 707 if (len1 > len2) { 708 res = memcmp(s1, s2, len2); 709 if (!res) 710 res = +1; 711 } else if (len1 < len2) { 712 res = memcmp(s1, s2, len1); 713 if (!res) 714 res = -1; 715 } else 716 res = memcmp(s1, s2, len1); 717 718 return (res); 719 720 } else { 721 int res = 0; 722 size_t i, maxlen; 723 724 i = 0; 725 maxlen = len1; 726 727 if (maxlen > len2) 728 maxlen = len2; 729 730 while (i < maxlen) { 731 /* goto next non-zero part: */ 732 while ((i < maxlen) && 733 !s1[i] && !s2[i]) 734 ++i; 735 736 if (i >= maxlen) 737 break; 738 739 if (s1[i] == 0) { 740 if (s2[i] == 0) 741 /* NOTREACHED */ 742 err(2, "bwscoll error 01"); 743 else 744 return (-1); 745 } else if (s2[i] == 0) 746 return (+1); 747 748 res = strcoll((const char*)(s1 + i), (const char*)(s2 + i)); 749 if (res) 750 return (res); 751 752 while ((i < maxlen) && 753 s1[i] && s2[i]) 754 ++i; 755 756 if (i >= maxlen) 757 break; 758 759 if (s1[i] == 0) { 760 if (s2[i] == 0) { 761 ++i; 762 continue; 763 } else 764 return (-1); 765 } else if (s2[i] == 0) 766 return (+1); 767 else 768 /* NOTREACHED */ 769 err(2, "bwscoll error 02"); 770 } 771 772 if (len1 < len2) 773 return (-1); 774 else if (len1 > len2) 775 return (+1); 776 777 return (0); 778 } 779 } else { 780 const wchar_t *s1, *s2; 781 size_t i, maxlen; 782 int res = 0; 783 784 s1 = bws1->data.wstr + offset; 785 s2 = bws2->data.wstr + offset; 786 787 i = 0; 788 maxlen = len1; 789 790 if (maxlen > len2) 791 maxlen = len2; 792 793 while (i < maxlen) { 794 795 /* goto next non-zero part: */ 796 while ((i < maxlen) && 797 !s1[i] && !s2[i]) 798 ++i; 799 800 if (i >= maxlen) 801 break; 802 803 if (s1[i] == 0) { 804 if (s2[i] == 0) 805 /* NOTREACHED */ 806 err(2, "bwscoll error 1"); 807 else 808 return (-1); 809 } else if (s2[i] == 0) 810 return (+1); 811 812 res = wide_str_coll(s1 + i, s2 + i); 813 if (res) 814 return (res); 815 816 while ((i < maxlen) && s1[i] && s2[i]) 817 ++i; 818 819 if (i >= maxlen) 820 break; 821 822 if (s1[i] == 0) { 823 if (s2[i] == 0) { 824 ++i; 825 continue; 826 } else 827 return (-1); 828 } else if (s2[i] == 0) 829 return (+1); 830 else 831 /* NOTREACHED */ 832 err(2, "bwscoll error 2"); 833 } 834 835 if (len1 < len2) 836 return (-1); 837 else if (len1 > len2) 838 return (+1); 839 840 return (0); 841 } 842 } 843 } 844 } 845 846 /* 847 * Correction of the system API 848 */ 849 double 850 bwstod(struct bwstring *s0, bool *empty) 851 { 852 double ret = 0; 853 854 if (MB_CUR_MAX == 1) { 855 unsigned char *end, *s; 856 char *ep; 857 858 s = s0->data.cstr; 859 end = s + s0->len; 860 ep = NULL; 861 862 while (isblank(*s) && s < end) 863 ++s; 864 865 if (!isprint(*s)) { 866 *empty = true; 867 return (0); 868 } 869 870 ret = strtod((char*)s, &ep); 871 if ((unsigned char*) ep == s) { 872 *empty = true; 873 return (0); 874 } 875 } else { 876 wchar_t *end, *ep, *s; 877 878 s = s0->data.wstr; 879 end = s + s0->len; 880 ep = NULL; 881 882 while (iswblank(*s) && s < end) 883 ++s; 884 885 if (!iswprint(*s)) { 886 *empty = true; 887 return (0); 888 } 889 890 ret = wcstod(s, &ep); 891 if (ep == s) { 892 *empty = true; 893 return (0); 894 } 895 } 896 897 *empty = false; 898 return (ret); 899 } 900 901 /* 902 * A helper function for monthcoll. If a line matches 903 * a month name, it returns (number of the month - 1), 904 * while if there is no match, it just return -1. 905 */ 906 907 int 908 bws_month_score(const struct bwstring *s0) 909 { 910 911 if (MB_CUR_MAX == 1) { 912 const unsigned char *end, *s; 913 914 s = s0->data.cstr; 915 end = s + s0->len; 916 917 while (isblank(*s) && s < end) 918 ++s; 919 920 for (int i = 11; i >= 0; --i) { 921 if (cmonths[i] && 922 (s == (unsigned char*)strstr((const char*)s, (char*)(cmonths[i])))) 923 return (i); 924 } 925 926 } else { 927 const wchar_t *end, *s; 928 929 s = s0->data.wstr; 930 end = s + s0->len; 931 932 while (iswblank(*s) && s < end) 933 ++s; 934 935 for (int i = 11; i >= 0; --i) { 936 if (wmonths[i] && (s == wcsstr(s, wmonths[i]))) 937 return (i); 938 } 939 } 940 941 return (-1); 942 } 943 944 /* 945 * Rips out leading blanks (-b). 946 */ 947 struct bwstring * 948 ignore_leading_blanks(struct bwstring *str) 949 { 950 951 if (MB_CUR_MAX == 1) { 952 unsigned char *dst, *end, *src; 953 954 src = str->data.cstr; 955 dst = src; 956 end = src + str->len; 957 958 while (src < end && isblank(*src)) 959 ++src; 960 961 if (src != dst) { 962 size_t newlen; 963 964 newlen = BWSLEN(str) - (src - dst); 965 966 while (src < end) { 967 *dst = *src; 968 ++dst; 969 ++src; 970 } 971 bws_setlen(str, newlen); 972 } 973 } else { 974 wchar_t *dst, *end, *src; 975 976 src = str->data.wstr; 977 dst = src; 978 end = src + str->len; 979 980 while (src < end && iswblank(*src)) 981 ++src; 982 983 if (src != dst) { 984 985 size_t newlen = BWSLEN(str) - (src - dst); 986 987 while (src < end) { 988 *dst = *src; 989 ++dst; 990 ++src; 991 } 992 bws_setlen(str, newlen); 993 994 } 995 } 996 return (str); 997 } 998 999 /* 1000 * Rips out nonprinting characters (-i). 1001 */ 1002 struct bwstring * 1003 ignore_nonprinting(struct bwstring *str) 1004 { 1005 size_t newlen = str->len; 1006 1007 if (MB_CUR_MAX == 1) { 1008 unsigned char *dst, *end, *src; 1009 unsigned char c; 1010 1011 src = str->data.cstr; 1012 dst = src; 1013 end = src + str->len; 1014 1015 while (src < end) { 1016 c = *src; 1017 if (isprint(c)) { 1018 *dst = c; 1019 ++dst; 1020 ++src; 1021 } else { 1022 ++src; 1023 --newlen; 1024 } 1025 } 1026 } else { 1027 wchar_t *dst, *end, *src; 1028 wchar_t c; 1029 1030 src = str->data.wstr; 1031 dst = src; 1032 end = src + str->len; 1033 1034 while (src < end) { 1035 c = *src; 1036 if (iswprint(c)) { 1037 *dst = c; 1038 ++dst; 1039 ++src; 1040 } else { 1041 ++src; 1042 --newlen; 1043 } 1044 } 1045 } 1046 bws_setlen(str, newlen); 1047 1048 return (str); 1049 } 1050 1051 /* 1052 * Rips out any characters that are not alphanumeric characters 1053 * nor blanks (-d). 1054 */ 1055 struct bwstring * 1056 dictionary_order(struct bwstring *str) 1057 { 1058 size_t newlen = str->len; 1059 1060 if (MB_CUR_MAX == 1) { 1061 unsigned char *dst, *end, *src; 1062 unsigned char c; 1063 1064 src = str->data.cstr; 1065 dst = src; 1066 end = src + str->len; 1067 1068 while (src < end) { 1069 c = *src; 1070 if (isalnum(c) || isblank(c)) { 1071 *dst = c; 1072 ++dst; 1073 ++src; 1074 } else { 1075 ++src; 1076 --newlen; 1077 } 1078 } 1079 } else { 1080 wchar_t *dst, *end, *src; 1081 wchar_t c; 1082 1083 src = str->data.wstr; 1084 dst = src; 1085 end = src + str->len; 1086 1087 while (src < end) { 1088 c = *src; 1089 if (iswalnum(c) || iswblank(c)) { 1090 *dst = c; 1091 ++dst; 1092 ++src; 1093 } else { 1094 ++src; 1095 --newlen; 1096 } 1097 } 1098 } 1099 bws_setlen(str, newlen); 1100 1101 return (str); 1102 } 1103 1104 /* 1105 * Converts string to lower case(-f). 1106 */ 1107 struct bwstring * 1108 ignore_case(struct bwstring *str) 1109 { 1110 1111 if (MB_CUR_MAX == 1) { 1112 unsigned char *end, *s; 1113 1114 s = str->data.cstr; 1115 end = s + str->len; 1116 1117 while (s < end) { 1118 *s = toupper(*s); 1119 ++s; 1120 } 1121 } else { 1122 wchar_t *end, *s; 1123 1124 s = str->data.wstr; 1125 end = s + str->len; 1126 1127 while (s < end) { 1128 *s = towupper(*s); 1129 ++s; 1130 } 1131 } 1132 return (str); 1133 } 1134 1135 void 1136 bws_disorder_warnx(struct bwstring *s, const char *fn, size_t pos) 1137 { 1138 1139 if (MB_CUR_MAX == 1) 1140 warnx("%s:%zu: disorder: %s", fn, pos + 1, s->data.cstr); 1141 else 1142 warnx("%s:%zu: disorder: %ls", fn, pos + 1, s->data.wstr); 1143 } 1144