DXR is a code search and navigation tool aimed at making sense of large projects. It supports full-text and regex searches as well as structural queries.

Mercurial (b6d82b1a6b02)

VCS Links

Line Code
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171
// © 2016 and later: Unicode, Inc. and others.
// License & terms of use: http://www.unicode.org/copyright.html
/*
*******************************************************************************
*
*   Copyright (C) 1999-2014, International Business Machines
*   Corporation and others.  All Rights Reserved.
*
*******************************************************************************
*   file name:  uniset_props.cpp
*   encoding:   UTF-8
*   tab size:   8 (not used)
*   indentation:4
*
*   created on: 2004aug25
*   created by: Markus W. Scherer
*
*   Character property dependent functions moved here from uniset.cpp
*/

#include "unicode/utypes.h"
#include "unicode/uniset.h"
#include "unicode/parsepos.h"
#include "unicode/uchar.h"
#include "unicode/uscript.h"
#include "unicode/symtable.h"
#include "unicode/uset.h"
#include "unicode/locid.h"
#include "unicode/brkiter.h"
#include "uset_imp.h"
#include "ruleiter.h"
#include "cmemory.h"
#include "ucln_cmn.h"
#include "util.h"
#include "uvector.h"
#include "uprops.h"
#include "propname.h"
#include "normalizer2impl.h"
#include "uinvchar.h"
#include "uprops.h"
#include "charstr.h"
#include "cstring.h"
#include "mutex.h"
#include "umutex.h"
#include "uassert.h"
#include "hash.h"

U_NAMESPACE_USE

// Define UChar constants using hex for EBCDIC compatibility
// Used #define to reduce private static exports and memory access time.
#define SET_OPEN        ((UChar)0x005B) /*[*/
#define SET_CLOSE       ((UChar)0x005D) /*]*/
#define HYPHEN          ((UChar)0x002D) /*-*/
#define COMPLEMENT      ((UChar)0x005E) /*^*/
#define COLON           ((UChar)0x003A) /*:*/
#define BACKSLASH       ((UChar)0x005C) /*\*/
#define INTERSECTION    ((UChar)0x0026) /*&*/
#define UPPER_U         ((UChar)0x0055) /*U*/
#define LOWER_U         ((UChar)0x0075) /*u*/
#define OPEN_BRACE      ((UChar)123)    /*{*/
#define CLOSE_BRACE     ((UChar)125)    /*}*/
#define UPPER_P         ((UChar)0x0050) /*P*/
#define LOWER_P         ((UChar)0x0070) /*p*/
#define UPPER_N         ((UChar)78)     /*N*/
#define EQUALS          ((UChar)0x003D) /*=*/

//static const UChar POSIX_OPEN[]  = { SET_OPEN,COLON,0 };  // "[:"
static const UChar POSIX_CLOSE[] = { COLON,SET_CLOSE,0 };  // ":]"
//static const UChar PERL_OPEN[]   = { BACKSLASH,LOWER_P,0 }; // "\\p"
//static const UChar PERL_CLOSE[]  = { CLOSE_BRACE,0 };    // "}"
//static const UChar NAME_OPEN[]   = { BACKSLASH,UPPER_N,0 };  // "\\N"
static const UChar HYPHEN_RIGHT_BRACE[] = {HYPHEN,SET_CLOSE,0}; /*-]*/

// Special property set IDs
static const char ANY[]   = "ANY";   // [\u0000-\U0010FFFF]
static const char ASCII[] = "ASCII"; // [\u0000-\u007F]
static const char ASSIGNED[] = "Assigned"; // [:^Cn:]

// Unicode name property alias
#define NAME_PROP "na"
#define NAME_PROP_LENGTH 2

/**
 * Delimiter string used in patterns to close a category reference:
 * ":]".  Example: "[:Lu:]".
 */
//static const UChar CATEGORY_CLOSE[] = {COLON, SET_CLOSE, 0x0000}; /* ":]" */

// Cached sets ------------------------------------------------------------- ***

U_CDECL_BEGIN
static UBool U_CALLCONV uset_cleanup();

static UnicodeSet *uni32Singleton;
static icu::UInitOnce uni32InitOnce = U_INITONCE_INITIALIZER;

/**
 * Cleanup function for UnicodeSet
 */
static UBool U_CALLCONV uset_cleanup(void) {
    delete uni32Singleton;
    uni32Singleton = NULL;
    uni32InitOnce.reset();
    return TRUE;
}

U_CDECL_END

U_NAMESPACE_BEGIN

namespace {

// Cache some sets for other services -------------------------------------- ***
void U_CALLCONV createUni32Set(UErrorCode &errorCode) {
    U_ASSERT(uni32Singleton == NULL);
    uni32Singleton = new UnicodeSet(UNICODE_STRING_SIMPLE("[:age=3.2:]"), errorCode);
    if(uni32Singleton==NULL) {
        errorCode=U_MEMORY_ALLOCATION_ERROR;
    } else {
        uni32Singleton->freeze();
    }
    ucln_common_registerCleanup(UCLN_COMMON_USET, uset_cleanup);
}


U_CFUNC UnicodeSet *
uniset_getUnicode32Instance(UErrorCode &errorCode) {
    umtx_initOnce(uni32InitOnce, &createUni32Set, errorCode);
    return uni32Singleton;
}

// helper functions for matching of pattern syntax pieces ------------------ ***
// these functions are parallel to the PERL_OPEN etc. strings above

// using these functions is not only faster than UnicodeString::compare() and
// caseCompare(), but they also make UnicodeSet work for simple patterns when
// no Unicode properties data is available - when caseCompare() fails

static inline UBool
isPerlOpen(const UnicodeString &pattern, int32_t pos) {
    UChar c;
    return pattern.charAt(pos)==BACKSLASH && ((c=pattern.charAt(pos+1))==LOWER_P || c==UPPER_P);
}

/*static inline UBool
isPerlClose(const UnicodeString &pattern, int32_t pos) {
    return pattern.charAt(pos)==CLOSE_BRACE;
}*/

static inline UBool
isNameOpen(const UnicodeString &pattern, int32_t pos) {
    return pattern.charAt(pos)==BACKSLASH && pattern.charAt(pos+1)==UPPER_N;
}

static inline UBool
isPOSIXOpen(const UnicodeString &pattern, int32_t pos) {
    return pattern.charAt(pos)==SET_OPEN && pattern.charAt(pos+1)==COLON;
}

/*static inline UBool
isPOSIXClose(const UnicodeString &pattern, int32_t pos) {
    return pattern.charAt(pos)==COLON && pattern.charAt(pos+1)==SET_CLOSE;
}*/

// TODO memory debugging provided inside uniset.cpp
// could be made available here but probably obsolete with use of modern
// memory leak checker tools
#define _dbgct(me)

}  // namespace

//----------------------------------------------------------------
// Constructors &c
//----------------------------------------------------------------

/**
 * Constructs a set from the given pattern, optionally ignoring
 * white space.  See the class description for the syntax of the
 * pattern language.
 * @param pattern a string specifying what characters are in the set
 */
UnicodeSet::UnicodeSet(const UnicodeString& pattern,
                       UErrorCode& status) {
    applyPattern(pattern, status);
    _dbgct(this);
}

//----------------------------------------------------------------
// Public API
//----------------------------------------------------------------

UnicodeSet& UnicodeSet::applyPattern(const UnicodeString& pattern,
                                     UErrorCode& status) {
    // Equivalent to
    //   return applyPattern(pattern, USET_IGNORE_SPACE, NULL, status);
    // but without dependency on closeOver().
    ParsePosition pos(0);
    applyPatternIgnoreSpace(pattern, pos, NULL, status);
    if (U_FAILURE(status)) return *this;

    int32_t i = pos.getIndex();
    // Skip over trailing whitespace
    ICU_Utility::skipWhitespace(pattern, i, TRUE);
    if (i != pattern.length()) {
        status = U_ILLEGAL_ARGUMENT_ERROR;
    }
    return *this;
}

void
UnicodeSet::applyPatternIgnoreSpace(const UnicodeString& pattern,
                                    ParsePosition& pos,
                                    const SymbolTable* symbols,
                                    UErrorCode& status) {
    if (U_FAILURE(status)) {
        return;
    }
    if (isFrozen()) {
        status = U_NO_WRITE_PERMISSION;
        return;
    }
    // Need to build the pattern in a temporary string because
    // _applyPattern calls add() etc., which set pat to empty.
    UnicodeString rebuiltPat;
    RuleCharacterIterator chars(pattern, symbols, pos);
    applyPattern(chars, symbols, rebuiltPat, USET_IGNORE_SPACE, NULL, 0, status);
    if (U_FAILURE(status)) return;
    if (chars.inVariable()) {
        // syntaxError(chars, "Extra chars in variable value");
        status = U_MALFORMED_SET;
        return;
    }
    setPattern(rebuiltPat);
}

/**
 * Return true if the given position, in the given pattern, appears
 * to be the start of a UnicodeSet pattern.
 */
UBool UnicodeSet::resemblesPattern(const UnicodeString& pattern, int32_t pos) {
    return ((pos+1) < pattern.length() &&
            pattern.charAt(pos) == (UChar)91/*[*/) ||
        resemblesPropertyPattern(pattern, pos);
}

//----------------------------------------------------------------
// Implementation: Pattern parsing
//----------------------------------------------------------------

namespace {

/**
 * A small all-inline class to manage a UnicodeSet pointer.  Add
 * operator->() etc. as needed.
 */
class UnicodeSetPointer {
    UnicodeSet* p;
public:
    inline UnicodeSetPointer() : p(0) {}
    inline ~UnicodeSetPointer() { delete p; }
    inline UnicodeSet* pointer() { return p; }
    inline UBool allocate() {
        if (p == 0) {
            p = new UnicodeSet();
        }
        return p != 0;
    }
};

constexpr int32_t MAX_DEPTH = 100;

}  // namespace

/**
 * Parse the pattern from the given RuleCharacterIterator.  The
 * iterator is advanced over the parsed pattern.
 * @param chars iterator over the pattern characters.  Upon return
 * it will be advanced to the first character after the parsed
 * pattern, or the end of the iteration if all characters are
 * parsed.
 * @param symbols symbol table to use to parse and dereference
 * variables, or null if none.
 * @param rebuiltPat the pattern that was parsed, rebuilt or
 * copied from the input pattern, as appropriate.
 * @param options a bit mask of zero or more of the following:
 * IGNORE_SPACE, CASE.
 */
void UnicodeSet::applyPattern(RuleCharacterIterator& chars,
                              const SymbolTable* symbols,
                              UnicodeString& rebuiltPat,
                              uint32_t options,
                              UnicodeSet& (UnicodeSet::*caseClosure)(int32_t attribute),
                              int32_t depth,
                              UErrorCode& ec) {
    if (U_FAILURE(ec)) return;
    if (depth > MAX_DEPTH) {
        ec = U_ILLEGAL_ARGUMENT_ERROR;
        return;
    }

    // Syntax characters: [ ] ^ - & { }

    // Recognized special forms for chars, sets: c-c s-s s&s

    int32_t opts = RuleCharacterIterator::PARSE_VARIABLES |
                   RuleCharacterIterator::PARSE_ESCAPES;
    if ((options & USET_IGNORE_SPACE) != 0) {
        opts |= RuleCharacterIterator::SKIP_WHITESPACE;
    }

    UnicodeString patLocal, buf;
    UBool usePat = FALSE;
    UnicodeSetPointer scratch;
    RuleCharacterIterator::Pos backup;

    // mode: 0=before [, 1=between [...], 2=after ]
    // lastItem: 0=none, 1=char, 2=set
    int8_t lastItem = 0, mode = 0;
    UChar32 lastChar = 0;
    UChar op = 0;

    UBool invert = FALSE;

    clear();

    while (mode != 2 && !chars.atEnd()) {
        U_ASSERT((lastItem == 0 && op == 0) ||
                 (lastItem == 1 && (op == 0 || op == HYPHEN /*'-'*/)) ||
                 (lastItem == 2 && (op == 0 || op == HYPHEN /*'-'*/ ||
                                    op == INTERSECTION /*'&'*/)));

        UChar32 c = 0;
        UBool literal = FALSE;
        UnicodeSet* nested = 0; // alias - do not delete

        // -------- Check for property pattern

        // setMode: 0=none, 1=unicodeset, 2=propertypat, 3=preparsed
        int8_t setMode = 0;
        if (resemblesPropertyPattern(chars, opts)) {
            setMode = 2;
        }

        // -------- Parse '[' of opening delimiter OR nested set.
        // If there is a nested set, use `setMode' to define how
        // the set should be parsed.  If the '[' is part of the
        // opening delimiter for this pattern, parse special
        // strings "[", "[^", "[-", and "[^-".  Check for stand-in
        // characters representing a nested set in the symbol
        // table.

        else {
            // Prepare to backup if necessary
            chars.getPos(backup);
            c = chars.next(opts, literal, ec);
            if (U_FAILURE(ec)) return;

            if (c == 0x5B /*'['*/ && !literal) {
                if (mode == 1) {
                    chars.setPos(backup); // backup
                    setMode = 1;
                } else {
                    // Handle opening '[' delimiter
                    mode = 1;
                    patLocal.append((UChar) 0x5B /*'['*/);
                    chars.getPos(backup); // prepare to backup
                    c = chars.next(opts, literal, ec); 
                    if (U_FAILURE(ec)) return;
                    if (c == 0x5E /*'^'*/ && !literal) {
                        invert = TRUE;
                        patLocal.append((UChar) 0x5E /*'^'*/);
                        chars.getPos(backup); // prepare to backup
                        c = chars.next(opts, literal, ec);
                        if (U_FAILURE(ec)) return;
                    }
                    // Fall through to handle special leading '-';
                    // otherwise restart loop for nested [], \p{}, etc.
                    if (c == HYPHEN /*'-'*/) {
                        literal = TRUE;
                        // Fall through to handle literal '-' below
                    } else {
                        chars.setPos(backup); // backup
                        continue;
                    }
                }
            } else if (symbols != 0) {
                const UnicodeFunctor *m = symbols->lookupMatcher(c);
                if (m != 0) {
                    const UnicodeSet *ms = dynamic_cast<const UnicodeSet *>(m);
                    if (ms == NULL) {
                        ec = U_MALFORMED_SET;
                        return;
                    }
                    // casting away const, but `nested' won't be modified
                    // (important not to modify stored set)
                    nested = const_cast<UnicodeSet*>(ms);
                    setMode = 3;
                }
            }
        }

        // -------- Handle a nested set.  This either is inline in
        // the pattern or represented by a stand-in that has
        // previously been parsed and was looked up in the symbol
        // table.

        if (setMode != 0) {
            if (lastItem == 1) {
                if (op != 0) {
                    // syntaxError(chars, "Char expected after operator");
                    ec = U_MALFORMED_SET;
                    return;
                }
                add(lastChar, lastChar);
                _appendToPat(patLocal, lastChar, FALSE);
                lastItem = 0;
                op = 0;
            }

            if (op == HYPHEN /*'-'*/ || op == INTERSECTION /*'&'*/) {
                patLocal.append(op);
            }

            if (nested == 0) {
                // lazy allocation
                if (!scratch.allocate()) {
                    ec = U_MEMORY_ALLOCATION_ERROR;
                    return;
                }
                nested = scratch.pointer();
            }
            switch (setMode) {
            case 1:
                nested->applyPattern(chars, symbols, patLocal, options, caseClosure, depth + 1, ec);
                break;
            case 2:
                chars.skipIgnored(opts);
                nested->applyPropertyPattern(chars, patLocal, ec);
                if (U_FAILURE(ec)) return;
                break;
            case 3: // `nested' already parsed
                nested->_toPattern(patLocal, FALSE);
                break;
            }

            usePat = TRUE;

            if (mode == 0) {
                // Entire pattern is a category; leave parse loop
                *this = *nested;
                mode = 2;
                break;
            }

            switch (op) {
            case HYPHEN: /*'-'*/
                removeAll(*nested);
                break;
            case INTERSECTION: /*'&'*/
                retainAll(*nested);
                break;
            case 0:
                addAll(*nested);
                break;
            }

            op = 0;
            lastItem = 2;

            continue;
        }

        if (mode == 0) {
            // syntaxError(chars, "Missing '['");
            ec = U_MALFORMED_SET;
            return;
        }

        // -------- Parse special (syntax) characters.  If the
        // current character is not special, or if it is escaped,
        // then fall through and handle it below.

        if (!literal) {
            switch (c) {
            case 0x5D /*']'*/:
                if (lastItem == 1) {
                    add(lastChar, lastChar);
                    _appendToPat(patLocal, lastChar, FALSE);
                }
                // Treat final trailing '-' as a literal
                if (op == HYPHEN /*'-'*/) {
                    add(op, op);
                    patLocal.append(op);
                } else if (op == INTERSECTION /*'&'*/) {
                    // syntaxError(chars, "Trailing '&'");
                    ec = U_MALFORMED_SET;
                    return;
                }
                patLocal.append((UChar) 0x5D /*']'*/);
                mode = 2;
                continue;
            case HYPHEN /*'-'*/:
                if (op == 0) {
                    if (lastItem != 0) {
                        op = (UChar) c;
                        continue;
                    } else {
                        // Treat final trailing '-' as a literal
                        add(c, c);
                        c = chars.next(opts, literal, ec);
                        if (U_FAILURE(ec)) return;
                        if (c == 0x5D /*']'*/ && !literal) {
                            patLocal.append(HYPHEN_RIGHT_BRACE, 2);
                            mode = 2;
                            continue;
                        }
                    }
                }
                // syntaxError(chars, "'-' not after char or set");
                ec = U_MALFORMED_SET;
                return;
            case INTERSECTION /*'&'*/:
                if (lastItem == 2 && op == 0) {
                    op = (UChar) c;
                    continue;
                }
                // syntaxError(chars, "'&' not after set");
                ec = U_MALFORMED_SET;
                return;
            case 0x5E /*'^'*/:
                // syntaxError(chars, "'^' not after '['");
                ec = U_MALFORMED_SET;
                return;
            case 0x7B /*'{'*/:
                if (op != 0) {
                    // syntaxError(chars, "Missing operand after operator");
                    ec = U_MALFORMED_SET;
                    return;
                }
                if (lastItem == 1) {
                    add(lastChar, lastChar);
                    _appendToPat(patLocal, lastChar, FALSE);
                }
                lastItem = 0;
                buf.truncate(0);
                {
                    UBool ok = FALSE;
                    while (!chars.atEnd()) {
                        c = chars.next(opts, literal, ec);
                        if (U_FAILURE(ec)) return;
                        if (c == 0x7D /*'}'*/ && !literal) {
                            ok = TRUE;
                            break;
                        }
                        buf.append(c);
                    }
                    if (buf.length() < 1 || !ok) {
                        // syntaxError(chars, "Invalid multicharacter string");
                        ec = U_MALFORMED_SET;
                        return;
                    }
                }
                // We have new string. Add it to set and continue;
                // we don't need to drop through to the further
                // processing
                add(buf);
                patLocal.append((UChar) 0x7B /*'{'*/);
                _appendToPat(patLocal, buf, FALSE);
                patLocal.append((UChar) 0x7D /*'}'*/);
                continue;
            case SymbolTable::SYMBOL_REF:
                //         symbols  nosymbols
                // [a-$]   error    error (ambiguous)
                // [a$]    anchor   anchor
                // [a-$x]  var "x"* literal '$'
                // [a-$.]  error    literal '$'
                // *We won't get here in the case of var "x"
                {
                    chars.getPos(backup);
                    c = chars.next(opts, literal, ec);
                    if (U_FAILURE(ec)) return;
                    UBool anchor = (c == 0x5D /*']'*/ && !literal);
                    if (symbols == 0 && !anchor) {
                        c = SymbolTable::SYMBOL_REF;
                        chars.setPos(backup);
                        break; // literal '$'
                    }
                    if (anchor && op == 0) {
                        if (lastItem == 1) {
                            add(lastChar, lastChar);
                            _appendToPat(patLocal, lastChar, FALSE);
                        }
                        add(U_ETHER);
                        usePat = TRUE;
                        patLocal.append((UChar) SymbolTable::SYMBOL_REF);
                        patLocal.append((UChar) 0x5D /*']'*/);
                        mode = 2;
                        continue;
                    }
                    // syntaxError(chars, "Unquoted '$'");
                    ec = U_MALFORMED_SET;
                    return;
                }
            default:
                break;
            }
        }

        // -------- Parse literal characters.  This includes both
        // escaped chars ("\u4E01") and non-syntax characters
        // ("a").

        switch (lastItem) {
        case 0:
            lastItem = 1;
            lastChar = c;
            break;
        case 1:
            if (op == HYPHEN /*'-'*/) {
                if (lastChar >= c) {
                    // Don't allow redundant (a-a) or empty (b-a) ranges;
                    // these are most likely typos.
                    // syntaxError(chars, "Invalid range");
                    ec = U_MALFORMED_SET;
                    return;
                }
                add(lastChar, c);
                _appendToPat(patLocal, lastChar, FALSE);
                patLocal.append(op);
                _appendToPat(patLocal, c, FALSE);
                lastItem = 0;
                op = 0;
            } else {
                add(lastChar, lastChar);
                _appendToPat(patLocal, lastChar, FALSE);
                lastChar = c;
            }
            break;
        case 2:
            if (op != 0) {
                // syntaxError(chars, "Set expected after operator");
                ec = U_MALFORMED_SET;
                return;
            }
            lastChar = c;
            lastItem = 1;
            break;
        }
    }

    if (mode != 2) {
        // syntaxError(chars, "Missing ']'");
        ec = U_MALFORMED_SET;
        return;
    }

    chars.skipIgnored(opts);

    /**
     * Handle global flags (invert, case insensitivity).  If this
     * pattern should be compiled case-insensitive, then we need
     * to close over case BEFORE COMPLEMENTING.  This makes
     * patterns like /[^abc]/i work.
     */
    if ((options & USET_CASE_INSENSITIVE) != 0) {
        (this->*caseClosure)(USET_CASE_INSENSITIVE);
    }
    else if ((options & USET_ADD_CASE_MAPPINGS) != 0) {
        (this->*caseClosure)(USET_ADD_CASE_MAPPINGS);
    }
    if (invert) {
        complement();
    }

    // Use the rebuilt pattern (patLocal) only if necessary.  Prefer the
    // generated pattern.
    if (usePat) {
        rebuiltPat.append(patLocal);
    } else {
        _generatePattern(rebuiltPat, FALSE);
    }
    if (isBogus() && U_SUCCESS(ec)) {
        // We likely ran out of memory. AHHH!
        ec = U_MEMORY_ALLOCATION_ERROR;
    }
}

//----------------------------------------------------------------
// Property set implementation
//----------------------------------------------------------------

namespace {

static UBool numericValueFilter(UChar32 ch, void* context) {
    return u_getNumericValue(ch) == *(double*)context;
}

static UBool generalCategoryMaskFilter(UChar32 ch, void* context) {
    int32_t value = *(int32_t*)context;
    return (U_GET_GC_MASK((UChar32) ch) & value) != 0;
}

static UBool versionFilter(UChar32 ch, void* context) {
    static const UVersionInfo none = { 0, 0, 0, 0 };
    UVersionInfo v;
    u_charAge(ch, v);
    UVersionInfo* version = (UVersionInfo*)context;
    return uprv_memcmp(&v, &none, sizeof(v)) > 0 && uprv_memcmp(&v, version, sizeof(v)) <= 0;
}

typedef struct {
    UProperty prop;
    int32_t value;
} IntPropertyContext;

static UBool intPropertyFilter(UChar32 ch, void* context) {
    IntPropertyContext* c = (IntPropertyContext*)context;
    return u_getIntPropertyValue((UChar32) ch, c->prop) == c->value;
}

static UBool scriptExtensionsFilter(UChar32 ch, void* context) {
    return uscript_hasScript(ch, *(UScriptCode*)context);
}

}  // namespace

/**
 * Generic filter-based scanning code for UCD property UnicodeSets.
 */
void UnicodeSet::applyFilter(UnicodeSet::Filter filter,
                             void* context,
                             const UnicodeSet* inclusions,
                             UErrorCode &status) {
    if (U_FAILURE(status)) return;

    // Logically, walk through all Unicode characters, noting the start
    // and end of each range for which filter.contain(c) is
    // true.  Add each range to a set.
    //
    // To improve performance, use an inclusions set which
    // encodes information about character ranges that are known
    // to have identical properties.
    // inclusions contains the first characters of
    // same-value ranges for the given property.

    clear();

    UChar32 startHasProperty = -1;
    int32_t limitRange = inclusions->getRangeCount();

    for (int j=0; j<limitRange; ++j) {
        // get current range
        UChar32 start = inclusions->getRangeStart(j);
        UChar32 end = inclusions->getRangeEnd(j);

        // for all the code points in the range, process
        for (UChar32 ch = start; ch <= end; ++ch) {
            // only add to this UnicodeSet on inflection points --
            // where the hasProperty value changes to false
            if ((*filter)(ch, context)) {
                if (startHasProperty < 0) {
                    startHasProperty = ch;
                }
            } else if (startHasProperty >= 0) {
                add(startHasProperty, ch-1);
                startHasProperty = -1;
            }
        }
    }
    if (startHasProperty >= 0) {
        add((UChar32)startHasProperty, (UChar32)0x10FFFF);
    }
    if (isBogus() && U_SUCCESS(status)) {
        // We likely ran out of memory. AHHH!
        status = U_MEMORY_ALLOCATION_ERROR;
    }
}

namespace {

static UBool mungeCharName(char* dst, const char* src, int32_t dstCapacity) {
    /* Note: we use ' ' in compiler code page */
    int32_t j = 0;
    char ch;
    --dstCapacity; /* make room for term. zero */
    while ((ch = *src++) != 0) {
        if (ch == ' ' && (j==0 || (j>0 && dst[j-1]==' '))) {
            continue;
        }
        if (j >= dstCapacity) return FALSE;
        dst[j++] = ch;
    }
    if (j > 0 && dst[j-1] == ' ') --j;
    dst[j] = 0;
    return TRUE;
}

}  // namespace

//----------------------------------------------------------------
// Property set API
//----------------------------------------------------------------

#define FAIL(ec) {ec=U_ILLEGAL_ARGUMENT_ERROR; return *this;}

UnicodeSet&
UnicodeSet::applyIntPropertyValue(UProperty prop, int32_t value, UErrorCode& ec) {
    if (U_FAILURE(ec) || isFrozen()) { return *this; }
    if (prop == UCHAR_GENERAL_CATEGORY_MASK) {
        const UnicodeSet* inclusions = CharacterProperties::getInclusionsForProperty(prop, ec);
        applyFilter(generalCategoryMaskFilter, &value, inclusions, ec);
    } else if (prop == UCHAR_SCRIPT_EXTENSIONS) {
        const UnicodeSet* inclusions = CharacterProperties::getInclusionsForProperty(prop, ec);
        UScriptCode script = (UScriptCode)value;
        applyFilter(scriptExtensionsFilter, &script, inclusions, ec);
    } else if (0 <= prop && prop < UCHAR_BINARY_LIMIT) {
        if (value == 0 || value == 1) {
            const USet *set = u_getBinaryPropertySet(prop, &ec);
            if (U_FAILURE(ec)) { return *this; }
            copyFrom(*UnicodeSet::fromUSet(set), TRUE);
            if (value == 0) {
                complement();
            }
        } else {
            clear();
        }
    } else if (UCHAR_INT_START <= prop && prop < UCHAR_INT_LIMIT) {
        const UnicodeSet* inclusions = CharacterProperties::getInclusionsForProperty(prop, ec);
        IntPropertyContext c = {prop, value};
        applyFilter(intPropertyFilter, &c, inclusions, ec);
    } else {
        ec = U_ILLEGAL_ARGUMENT_ERROR;
    }
    return *this;
}

UnicodeSet&
UnicodeSet::applyPropertyAlias(const UnicodeString& prop,
                               const UnicodeString& value,
                               UErrorCode& ec) {
    if (U_FAILURE(ec) || isFrozen()) return *this;

    // prop and value used to be converted to char * using the default
    // converter instead of the invariant conversion.
    // This should not be necessary because all Unicode property and value
    // names use only invariant characters.
    // If there are any variant characters, then we won't find them anyway.
    // Checking first avoids assertion failures in the conversion.
    if( !uprv_isInvariantUString(prop.getBuffer(), prop.length()) ||
        !uprv_isInvariantUString(value.getBuffer(), value.length())
    ) {
        FAIL(ec);
    }
    CharString pname, vname;
    pname.appendInvariantChars(prop, ec);
    vname.appendInvariantChars(value, ec);
    if (U_FAILURE(ec)) return *this;

    UProperty p;
    int32_t v;
    UBool invert = FALSE;

    if (value.length() > 0) {
        p = u_getPropertyEnum(pname.data());
        if (p == UCHAR_INVALID_CODE) FAIL(ec);

        // Treat gc as gcm
        if (p == UCHAR_GENERAL_CATEGORY) {
            p = UCHAR_GENERAL_CATEGORY_MASK;
        }

        if ((p >= UCHAR_BINARY_START && p < UCHAR_BINARY_LIMIT) ||
            (p >= UCHAR_INT_START && p < UCHAR_INT_LIMIT) ||
            (p >= UCHAR_MASK_START && p < UCHAR_MASK_LIMIT)) {
            v = u_getPropertyValueEnum(p, vname.data());
            if (v == UCHAR_INVALID_CODE) {
                // Handle numeric CCC
                if (p == UCHAR_CANONICAL_COMBINING_CLASS ||
                    p == UCHAR_TRAIL_CANONICAL_COMBINING_CLASS ||
                    p == UCHAR_LEAD_CANONICAL_COMBINING_CLASS) {
                    char* end;
                    double val = uprv_strtod(vname.data(), &end);
                    // Anything between 0 and 255 is valid even if unused.
                    // Cast double->int only after range check.
                    // We catch NaN here because comparing it with both 0 and 255 will be false
                    // (as are all comparisons with NaN).
                    if (*end != 0 || !(0 <= val && val <= 255) ||
                            (v = (int32_t)val) != val) {
                        // non-integral value or outside 0..255, or trailing junk
                        FAIL(ec);
                    }
                } else {
                    FAIL(ec);
                }
            }
        }

        else {

            switch (p) {
            case UCHAR_NUMERIC_VALUE:
                {
                    char* end;
                    double val = uprv_strtod(vname.data(), &end);
                    if (*end != 0) {
                        FAIL(ec);
                    }
                    applyFilter(numericValueFilter, &val,
                                CharacterProperties::getInclusionsForProperty(p, ec), ec);
                    return *this;
                }
            case UCHAR_NAME:
                {
                    // Must munge name, since u_charFromName() does not do
                    // 'loose' matching.
                    char buf[128]; // it suffices that this be > uprv_getMaxCharNameLength
                    if (!mungeCharName(buf, vname.data(), sizeof(buf))) FAIL(ec);
                    UChar32 ch = u_charFromName(U_EXTENDED_CHAR_NAME, buf, &ec);
                    if (U_SUCCESS(ec)) {
                        clear();
                        add(ch);
                        return *this;
                    } else {
                        FAIL(ec);
                    }
                }
            case UCHAR_UNICODE_1_NAME:
                // ICU 49 deprecates the Unicode_1_Name property APIs.
                FAIL(ec);
            case UCHAR_AGE:
                {
                    // Must munge name, since u_versionFromString() does not do
                    // 'loose' matching.
                    char buf[128];
                    if (!mungeCharName(buf, vname.data(), sizeof(buf))) FAIL(ec);
                    UVersionInfo version;
                    u_versionFromString(version, buf);
                    applyFilter(versionFilter, &version,
                                CharacterProperties::getInclusionsForProperty(p, ec), ec);
                    return *this;
                }
            case UCHAR_SCRIPT_EXTENSIONS:
                v = u_getPropertyValueEnum(UCHAR_SCRIPT, vname.data());
                if (v == UCHAR_INVALID_CODE) {
                    FAIL(ec);
                }
                // fall through to calling applyIntPropertyValue()
                break;
            default:
                // p is a non-binary, non-enumerated property that we
                // don't support (yet).
                FAIL(ec);
            }
        }
    }

    else {
        // value is empty.  Interpret as General Category, Script, or
        // Binary property.
        p = UCHAR_GENERAL_CATEGORY_MASK;
        v = u_getPropertyValueEnum(p, pname.data());
        if (v == UCHAR_INVALID_CODE) {
            p = UCHAR_SCRIPT;
            v = u_getPropertyValueEnum(p, pname.data());
            if (v == UCHAR_INVALID_CODE) {
                p = u_getPropertyEnum(pname.data());
                if (p >= UCHAR_BINARY_START && p < UCHAR_BINARY_LIMIT) {
                    v = 1;
                } else if (0 == uprv_comparePropertyNames(ANY, pname.data())) {
                    set(MIN_VALUE, MAX_VALUE);
                    return *this;
                } else if (0 == uprv_comparePropertyNames(ASCII, pname.data())) {
                    set(0, 0x7F);
                    return *this;
                } else if (0 == uprv_comparePropertyNames(ASSIGNED, pname.data())) {
                    // [:Assigned:]=[:^Cn:]
                    p = UCHAR_GENERAL_CATEGORY_MASK;
                    v = U_GC_CN_MASK;
                    invert = TRUE;
                } else {
                    FAIL(ec);
                }
            }
        }
    }

    applyIntPropertyValue(p, v, ec);
    if(invert) {
        complement();
    }

    if (isBogus() && U_SUCCESS(ec)) {
        // We likely ran out of memory. AHHH!
        ec = U_MEMORY_ALLOCATION_ERROR;
    }
    return *this;
}

//----------------------------------------------------------------
// Property set patterns
//----------------------------------------------------------------

/**
 * Return true if the given position, in the given pattern, appears
 * to be the start of a property set pattern.
 */
UBool UnicodeSet::resemblesPropertyPattern(const UnicodeString& pattern,
                                           int32_t pos) {
    // Patterns are at least 5 characters long
    if ((pos+5) > pattern.length()) {
        return FALSE;
    }

    // Look for an opening [:, [:^, \p, or \P
    return isPOSIXOpen(pattern, pos) || isPerlOpen(pattern, pos) || isNameOpen(pattern, pos);
}

/**
 * Return true if the given iterator appears to point at a
 * property pattern.  Regardless of the result, return with the
 * iterator unchanged.
 * @param chars iterator over the pattern characters.  Upon return
 * it will be unchanged.
 * @param iterOpts RuleCharacterIterator options
 */
UBool UnicodeSet::resemblesPropertyPattern(RuleCharacterIterator& chars,
                                           int32_t iterOpts) {
    // NOTE: literal will always be FALSE, because we don't parse escapes.
    UBool result = FALSE, literal;
    UErrorCode ec = U_ZERO_ERROR;
    iterOpts &= ~RuleCharacterIterator::PARSE_ESCAPES;
    RuleCharacterIterator::Pos pos;
    chars.getPos(pos);
    UChar32 c = chars.next(iterOpts, literal, ec);
    if (c == 0x5B /*'['*/ || c == 0x5C /*'\\'*/) {
        UChar32 d = chars.next(iterOpts & ~RuleCharacterIterator::SKIP_WHITESPACE,
                               literal, ec);
        result = (c == 0x5B /*'['*/) ? (d == 0x3A /*':'*/) :
                 (d == 0x4E /*'N'*/ || d == 0x70 /*'p'*/ || d == 0x50 /*'P'*/);
    }
    chars.setPos(pos);
    return result && U_SUCCESS(ec);
}

/**
 * Parse the given property pattern at the given parse position.
 */
UnicodeSet& UnicodeSet::applyPropertyPattern(const UnicodeString& pattern,
                                             ParsePosition& ppos,
                                             UErrorCode &ec) {
    int32_t pos = ppos.getIndex();

    UBool posix = FALSE; // true for [:pat:], false for \p{pat} \P{pat} \N{pat}
    UBool isName = FALSE; // true for \N{pat}, o/w false
    UBool invert = FALSE;

    if (U_FAILURE(ec)) return *this;

    // Minimum length is 5 characters, e.g. \p{L}
    if ((pos+5) > pattern.length()) {
        FAIL(ec);
    }

    // On entry, ppos should point to one of the following locations:
    // Look for an opening [:, [:^, \p, or \P
    if (isPOSIXOpen(pattern, pos)) {
        posix = TRUE;
        pos += 2;
        pos = ICU_Utility::skipWhitespace(pattern, pos);
        if (pos < pattern.length() && pattern.charAt(pos) == COMPLEMENT) {
            ++pos;
            invert = TRUE;
        }
    } else if (isPerlOpen(pattern, pos) || isNameOpen(pattern, pos)) {
        UChar c = pattern.charAt(pos+1);
        invert = (c == UPPER_P);
        isName = (c == UPPER_N);
        pos += 2;
        pos = ICU_Utility::skipWhitespace(pattern, pos);
        if (pos == pattern.length() || pattern.charAt(pos++) != OPEN_BRACE) {
            // Syntax error; "\p" or "\P" not followed by "{"
            FAIL(ec);
        }
    } else {
        // Open delimiter not seen
        FAIL(ec);
    }

    // Look for the matching close delimiter, either :] or }
    int32_t close;
    if (posix) {
      close = pattern.indexOf(POSIX_CLOSE, 2, pos);
    } else {
      close = pattern.indexOf(CLOSE_BRACE, pos);
    }
    if (close < 0) {
        // Syntax error; close delimiter missing
        FAIL(ec);
    }

    // Look for an '=' sign.  If this is present, we will parse a
    // medium \p{gc=Cf} or long \p{GeneralCategory=Format}
    // pattern.
    int32_t equals = pattern.indexOf(EQUALS, pos);
    UnicodeString propName, valueName;
    if (equals >= 0 && equals < close && !isName) {
        // Equals seen; parse medium/long pattern
        pattern.extractBetween(pos, equals, propName);
        pattern.extractBetween(equals+1, close, valueName);
    }

    else {
        // Handle case where no '=' is seen, and \N{}
        pattern.extractBetween(pos, close, propName);
            
        // Handle \N{name}
        if (isName) {
            // This is a little inefficient since it means we have to
            // parse NAME_PROP back to UCHAR_NAME even though we already
            // know it's UCHAR_NAME.  If we refactor the API to
            // support args of (UProperty, char*) then we can remove
            // NAME_PROP and make this a little more efficient.
            valueName = propName;
            propName = UnicodeString(NAME_PROP, NAME_PROP_LENGTH, US_INV);
        }
    }

    applyPropertyAlias(propName, valueName, ec);

    if (U_SUCCESS(ec)) {
        if (invert) {
            complement();
        }
            
        // Move to the limit position after the close delimiter if the
        // parse succeeded.
        ppos.setIndex(close + (posix ? 2 : 1));
    }

    return *this;
}

/**
 * Parse a property pattern.
 * @param chars iterator over the pattern characters.  Upon return
 * it will be advanced to the first character after the parsed
 * pattern, or the end of the iteration if all characters are
 * parsed.
 * @param rebuiltPat the pattern that was parsed, rebuilt or
 * copied from the input pattern, as appropriate.
 */
void UnicodeSet::applyPropertyPattern(RuleCharacterIterator& chars,
                                      UnicodeString& rebuiltPat,
                                      UErrorCode& ec) {
    if (U_FAILURE(ec)) return;
    UnicodeString pattern;
    chars.lookahead(pattern);
    ParsePosition pos(0);
    applyPropertyPattern(pattern, pos, ec);
    if (U_FAILURE(ec)) return;
    if (pos.getIndex() == 0) {
        // syntaxError(chars, "Invalid property pattern");
        ec = U_MALFORMED_SET;
        return;
    }
    chars.jumpahead(pos.getIndex());
    rebuiltPat.append(pattern, 0, pos.getIndex());
}

U_NAMESPACE_END