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 (27a812186ff4)

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
/* -*- Mode: C++; tab-width: 8; indent-tabs-mode: nil; c-basic-offset: 4 -*-
 * vim: set ts=8 sts=4 et sw=4 tw=99:
 *
 * Copyright (C) 2009 Apple Inc. All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 *
 * THIS SOFTWARE IS PROVIDED BY APPLE INC. ``AS IS'' AND ANY
 * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
 * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL APPLE INC. OR
 * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
 * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
 * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
 * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
 * OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
 * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 
 */

#ifndef yarr_YarrParser_h
#define yarr_YarrParser_h

#include "yarr/Yarr.h"

namespace JSC { namespace Yarr {

enum BuiltInCharacterClassID {
    DigitClassID,
    SpaceClassID,
    WordClassID,
    NewlineClassID
};

// The Parser class should not be used directly - only via the Yarr::parse() method.
template<class Delegate, typename CharType>
class Parser {
private:
    template<class FriendDelegate>
    friend ErrorCode parse(FriendDelegate&, const String& pattern, unsigned backReferenceLimit);

    /*
     * CharacterClassParserDelegate:
     *
     * The class CharacterClassParserDelegate is used in the parsing of character
     * classes.  This class handles detection of character ranges.  This class
     * implements enough of the delegate interface such that it can be passed to
     * parseEscape() as an EscapeDelegate.  This allows parseEscape() to be reused
     * to perform the parsing of escape characters in character sets.
     */
    class CharacterClassParserDelegate {
    public:
        CharacterClassParserDelegate(Delegate& delegate, ErrorCode& err)
            : m_delegate(delegate)
            , m_err(err)
            , m_state(Empty)
            , m_character(0)
        {
        }

        /*
         * begin():
         *
         * Called at beginning of construction.
         */
        void begin(bool invert)
        {
            m_delegate.atomCharacterClassBegin(invert);
        }

        /*
         * atomPatternCharacter():
         *
         * This method is called either from parseCharacterClass() (for an unescaped
         * character in a character class), or from parseEscape(). In the former case
         * the value true will be passed for the argument 'hyphenIsRange', and in this
         * mode we will allow a hypen to be treated as indicating a range (i.e. /[a-z]/
         * is different to /[a\-z]/).
         */
        void atomPatternCharacter(UChar ch, bool hyphenIsRange = false)
        {
            switch (m_state) {
            case AfterCharacterClass:
                // Following a builtin character class we need look out for a hyphen.
                // We're looking for invalid ranges, such as /[\d-x]/ or /[\d-\d]/.
                // If we see a hyphen following a charater class then unlike usual
                // we'll report it to the delegate immediately, and put ourself into
                // a poisoned state. Any following calls to add another character or
                // character class will result in an error. (A hypen following a
                // character-class is itself valid, but only  at the end of a regex).
                if (hyphenIsRange && ch == '-') {
                    m_delegate.atomCharacterClassAtom('-');
                    m_state = AfterCharacterClassHyphen;
                    return;
                }
                // Otherwise just fall through - cached character so treat this as Empty.

            case Empty:
                m_character = ch;
                m_state = CachedCharacter;
                return;

            case CachedCharacter:
                if (hyphenIsRange && ch == '-')
                    m_state = CachedCharacterHyphen;
                else {
                    m_delegate.atomCharacterClassAtom(m_character);
                    m_character = ch;
                }
                return;

            case CachedCharacterHyphen:
                if (ch < m_character) {
                    m_err = CharacterClassOutOfOrder;
                    return;
                }
                m_delegate.atomCharacterClassRange(m_character, ch);
                m_state = Empty;
                return;

            case AfterCharacterClassHyphen:
                m_delegate.atomCharacterClassAtom(ch);
                m_state = Empty;
                return;
            }
        }

        /*
         * atomBuiltInCharacterClass():
         *
         * Adds a built-in character class, called by parseEscape().
         */
        void atomBuiltInCharacterClass(BuiltInCharacterClassID classID, bool invert)
        {
            switch (m_state) {
            case CachedCharacter:
                // Flush the currently cached character, then fall through.
                m_delegate.atomCharacterClassAtom(m_character);

            case Empty:
            case AfterCharacterClass:
                m_state = AfterCharacterClass;
                m_delegate.atomCharacterClassBuiltIn(classID, invert);
                return;

            case CachedCharacterHyphen:
                // Error! We have a range that looks like [x-\d]. We require
                // the end of the range to be a single character.
                m_err = CharacterClassInvalidRange;
                return;
            case AfterCharacterClassHyphen:
                m_delegate.atomCharacterClassBuiltIn(classID, invert);
                m_state = Empty;
                return;
            }
        }

        /*
         * end():
         *
         * Called at end of construction.
         */
        void end()
        {
            if (m_state == CachedCharacter)
                m_delegate.atomCharacterClassAtom(m_character);
            else if (m_state == CachedCharacterHyphen) {
                m_delegate.atomCharacterClassAtom(m_character);
                m_delegate.atomCharacterClassAtom('-');
            }
            m_delegate.atomCharacterClassEnd();
        }

        // parseEscape() should never call these delegate methods when
        // invoked with inCharacterClass set.
        NO_RETURN_DUE_TO_ASSERT void assertionWordBoundary(bool) { ASSERT_NOT_REACHED(); }
        NO_RETURN_DUE_TO_ASSERT void atomBackReference(unsigned) { ASSERT_NOT_REACHED(); }

    private:
        Delegate& m_delegate;
        ErrorCode& m_err;
        enum CharacterClassConstructionState {
            Empty,
            CachedCharacter,
            CachedCharacterHyphen,
            AfterCharacterClass,
            AfterCharacterClassHyphen
        } m_state;
        UChar m_character;
    };

    Parser(Delegate& delegate, const String& pattern, unsigned backReferenceLimit)
        : m_delegate(delegate)
        , m_backReferenceLimit(backReferenceLimit)
        , m_err(NoError)
        , m_data(pattern.chars())
        , m_size(pattern.length())
        , m_index(0)
        , m_parenthesesNestingDepth(0)
    {
    }

    /*
     * parseEscape():
     *
     * Helper for parseTokens() AND parseCharacterClass().
     * Unlike the other parser methods, this function does not report tokens
     * directly to the member delegate (m_delegate), instead tokens are
     * emitted to the delegate provided as an argument.  In the case of atom
     * escapes, parseTokens() will call parseEscape() passing m_delegate as
     * an argument, and as such the escape will be reported to the delegate.
     *
     * However this method may also be used by parseCharacterClass(), in which
     * case a CharacterClassParserDelegate will be passed as the delegate that
     * tokens should be added to.  A boolean flag is also provided to indicate
     * whether that an escape in a CharacterClass is being parsed (some parsing
     * rules change in this context).
     *
     * The boolean value returned by this method indicates whether the token
     * parsed was an atom (outside of a characted class \b and \B will be
     * interpreted as assertions).
     */
    template<bool inCharacterClass, class EscapeDelegate>
    bool parseEscape(EscapeDelegate& delegate)
    {
        ASSERT(!m_err);
        ASSERT(peek() == '\\');
        consume();

        if (atEndOfPattern()) {
            m_err = EscapeUnterminated;
            return false;
        }

        switch (peek()) {
        // Assertions
        case 'b':
            consume();
            if (inCharacterClass)
                delegate.atomPatternCharacter('\b');
            else {
                delegate.assertionWordBoundary(false);
                return false;
            }
            break;
        case 'B':
            consume();
            if (inCharacterClass)
                delegate.atomPatternCharacter('B');
            else {
                delegate.assertionWordBoundary(true);
                return false;
            }
            break;

        // CharacterClassEscape
        case 'd':
            consume();
            delegate.atomBuiltInCharacterClass(DigitClassID, false);
            break;
        case 's':
            consume();
            delegate.atomBuiltInCharacterClass(SpaceClassID, false);
            break;
        case 'w':
            consume();
            delegate.atomBuiltInCharacterClass(WordClassID, false);
            break;
        case 'D':
            consume();
            delegate.atomBuiltInCharacterClass(DigitClassID, true);
            break;
        case 'S':
            consume();
            delegate.atomBuiltInCharacterClass(SpaceClassID, true);
            break;
        case 'W':
            consume();
            delegate.atomBuiltInCharacterClass(WordClassID, true);
            break;

        // DecimalEscape
        case '1':
        case '2':
        case '3':
        case '4':
        case '5':
        case '6':
        case '7':
        case '8':
        case '9': {
            // To match Firefox, we parse an invalid backreference in the range [1-7] as an octal escape.
            // First, try to parse this as backreference.
            if (!inCharacterClass) {
                ParseState state = saveState();

                unsigned backReference;
                if (!consumeNumber(backReference))
                    break; 
                if (backReference <= m_backReferenceLimit) {
                    delegate.atomBackReference(backReference);
                    break;
                }

                restoreState(state);
            }
            
            // Not a backreference, and not octal.
            if (peek() >= '8') {
                delegate.atomPatternCharacter('\\');
                break;
            }

            // Fall-through to handle this as an octal escape.
        }

        // Octal escape
        case '0':
            delegate.atomPatternCharacter(consumeOctal());
            break;

        // ControlEscape
        case 'f':
            consume();
            delegate.atomPatternCharacter('\f');
            break;
        case 'n':
            consume();
            delegate.atomPatternCharacter('\n');
            break;
        case 'r':
            consume();
            delegate.atomPatternCharacter('\r');
            break;
        case 't':
            consume();
            delegate.atomPatternCharacter('\t');
            break;
        case 'v':
            consume();
            delegate.atomPatternCharacter('\v');
            break;

        // ControlLetter
        case 'c': {
            ParseState state = saveState();
            consume();
            if (!atEndOfPattern()) {
                int control = consume();

                // To match Firefox, inside a character class, we also accept numbers and '_' as control characters.
                if (inCharacterClass ? WTF::isASCIIAlphanumeric(control) || (control == '_') : WTF::isASCIIAlpha(control)) {
                    delegate.atomPatternCharacter(control & 0x1f);
                    break;
                }
            }
            restoreState(state);
            delegate.atomPatternCharacter('\\');
            break;
        }

        // HexEscape
        case 'x': {
            consume();
            int x = tryConsumeHex(2);
            if (x == -1)
                delegate.atomPatternCharacter('x');
            else
                delegate.atomPatternCharacter(x);
            break;
        }

        // UnicodeEscape
        case 'u': {
            consume();
            int u = tryConsumeHex(4);
            if (u == -1)
                delegate.atomPatternCharacter('u');
            else
                delegate.atomPatternCharacter(u);
            break;
        }

        // IdentityEscape
        default:
            delegate.atomPatternCharacter(consume());
        }
        
        return true;
    }

    /*
     * parseAtomEscape(), parseCharacterClassEscape():
     *
     * These methods alias to parseEscape().
     */
    bool parseAtomEscape()
    {
        return parseEscape<false>(m_delegate);
    }
    void parseCharacterClassEscape(CharacterClassParserDelegate& delegate)
    {
        parseEscape<true>(delegate);
    }

    /*
     * parseCharacterClass():
     *
     * Helper for parseTokens(); calls dirctly and indirectly (via parseCharacterClassEscape)
     * to an instance of CharacterClassParserDelegate, to describe the character class to the
     * delegate.
     */
    void parseCharacterClass()
    {
        ASSERT(!m_err);
        ASSERT(peek() == '[');
        consume();

        CharacterClassParserDelegate characterClassConstructor(m_delegate, m_err);

        characterClassConstructor.begin(tryConsume('^'));

        while (!atEndOfPattern()) {
            switch (peek()) {
            case ']':
                consume();
                characterClassConstructor.end();
                return;

            case '\\':
                parseCharacterClassEscape(characterClassConstructor);
                break;

            default:
                characterClassConstructor.atomPatternCharacter(consume(), true);
            }

            if (m_err)
                return;
        }

        m_err = CharacterClassUnmatched;
    }

    /*
     * parseParenthesesBegin():
     *
     * Helper for parseTokens(); checks for parentheses types other than regular capturing subpatterns.
     */
    void parseParenthesesBegin()
    {
        ASSERT(!m_err);
        ASSERT(peek() == '(');
        consume();

        if (tryConsume('?')) {
            if (atEndOfPattern()) {
                m_err = ParenthesesTypeInvalid;
                return;
            }

            switch (consume()) {
            case ':':
                m_delegate.atomParenthesesSubpatternBegin(false);
                break;
            
            case '=':
                m_delegate.atomParentheticalAssertionBegin();
                break;

            case '!':
                m_delegate.atomParentheticalAssertionBegin(true);
                break;
            
            default:
                m_err = ParenthesesTypeInvalid;
            }
        } else
            m_delegate.atomParenthesesSubpatternBegin();

        ++m_parenthesesNestingDepth;
    }

    /*
     * parseParenthesesEnd():
     *
     * Helper for parseTokens(); checks for parse errors (due to unmatched parentheses).
     */
    void parseParenthesesEnd()
    {
        ASSERT(!m_err);
        ASSERT(peek() == ')');
        consume();

        if (m_parenthesesNestingDepth > 0)
            m_delegate.atomParenthesesEnd();
        else
            m_err = ParenthesesUnmatched;

        --m_parenthesesNestingDepth;
    }

    /*
     * parseQuantifier():
     *
     * Helper for parseTokens(); checks for parse errors and non-greedy quantifiers.
     */
    void parseQuantifier(bool lastTokenWasAnAtom, unsigned min, unsigned max)
    {
        ASSERT(!m_err);
        ASSERT(min <= max);

        if (min == UINT_MAX) {
            m_err = QuantifierTooLarge;
            return;
        }

        if (lastTokenWasAnAtom)
            m_delegate.quantifyAtom(min, max, !tryConsume('?'));
        else
            m_err = QuantifierWithoutAtom;
    }

    /*
     * parseTokens():
     *
     * This method loops over the input pattern reporting tokens to the delegate.
     * The method returns when a parse error is detected, or the end of the pattern
     * is reached.  One piece of state is tracked around the loop, which is whether
     * the last token passed to the delegate was an atom (this is necessary to detect
     * a parse error when a quantifier provided without an atom to quantify).
     */
    void parseTokens()
    {
        bool lastTokenWasAnAtom = false;

        while (!atEndOfPattern()) {
            switch (peek()) {
            case '|':
                consume();
                m_delegate.disjunction();
                lastTokenWasAnAtom = false;
                break;

            case '(':
                parseParenthesesBegin();
                lastTokenWasAnAtom = false;
                break;

            case ')':
                parseParenthesesEnd();
                lastTokenWasAnAtom = true;
                break;

            case '^':
                consume();
                m_delegate.assertionBOL();
                lastTokenWasAnAtom = false;
                break;

            case '$':
                consume();
                m_delegate.assertionEOL();
                lastTokenWasAnAtom = false;
                break;

            case '.':
                consume();
                m_delegate.atomBuiltInCharacterClass(NewlineClassID, true);
                lastTokenWasAnAtom = true;
                break;

            case '[':
                parseCharacterClass();
                lastTokenWasAnAtom = true;
                break;

            case '\\':
                lastTokenWasAnAtom = parseAtomEscape();
                break;

            case '*':
                consume();
                parseQuantifier(lastTokenWasAnAtom, 0, quantifyInfinite);
                lastTokenWasAnAtom = false;
                break;

            case '+':
                consume();
                parseQuantifier(lastTokenWasAnAtom, 1, quantifyInfinite);
                lastTokenWasAnAtom = false;
                break;

            case '?':
                consume();
                parseQuantifier(lastTokenWasAnAtom, 0, 1);
                lastTokenWasAnAtom = false;
                break;

            case '{': {
                ParseState state = saveState();

                consume();
                if (peekIsDigit()) {
                    unsigned min;
                    if (!consumeNumber(min))
                        break;
                    // Clamping to INT_MAX is technically a spec deviation. In practice, it's
                    // undetectable, because we can't even allocate strings large enough for
                    // quantifiers this large to ever create different results than smaller ones.
                    if (min > INT_MAX)
                        min = INT_MAX;

                    unsigned max = min;
                    if (tryConsume(',')) {
                        if (peekIsDigit()) {
                            if (!consumeNumber(max))
                                break;
                            // Clamping to INT_MAX is technically a spec deviation. In practice,
                            // it's undetectable, because we can't even allocate strings large
                            // enough for quantifiers this large to ever create different results
                            // than smaller ones.
                            if (max > INT_MAX)
                                max = INT_MAX;
                        } else {
                            max = quantifyInfinite;
                        }
                    }

                    if (tryConsume('}')) {
                        if (min <= max)
                            parseQuantifier(lastTokenWasAnAtom, min, max);
                        else
                            m_err = QuantifierOutOfOrder;
                        lastTokenWasAnAtom = false;
                        break;
                    }
                }

                restoreState(state);
            } // if we did not find a complete quantifer, fall through to the default case.

            default:
                m_delegate.atomPatternCharacter(consume());
                lastTokenWasAnAtom = true;
            }

            if (m_err)
                return;
        }

        if (m_parenthesesNestingDepth > 0)
            m_err = MissingParentheses;
    }

    /*
     * parse():
     *
     * This method calls parseTokens() to parse over the input and converts any
     * error code to a const char* for a result.
     */
    ErrorCode parse()
    {
        if (m_size > MAX_PATTERN_SIZE)
            m_err = PatternTooLarge;
        else
            parseTokens();
        ASSERT(atEndOfPattern() || m_err);

        return m_err;
    }

    // Misc helper functions:

    typedef unsigned ParseState;
    
    ParseState saveState()
    {
        return m_index;
    }

    void restoreState(ParseState state)
    {
        m_index = state;
    }

    bool atEndOfPattern()
    {
        ASSERT(m_index <= m_size);
        return m_index == m_size;
    }

    int peek()
    {
        ASSERT(m_index < m_size);
        return m_data[m_index];
    }

    bool peekIsDigit()
    {
        return !atEndOfPattern() && WTF::isASCIIDigit(peek());
    }

    unsigned peekDigit()
    {
        ASSERT(peekIsDigit());
        return peek() - '0';
    }

    int consume()
    {
        ASSERT(m_index < m_size);
        return m_data[m_index++];
    }

    unsigned consumeDigit()
    {
        ASSERT(peekIsDigit());
        return consume() - '0';
    }

    bool consumeNumber(unsigned& accum)
    {
        accum = consumeDigit();
        while (peekIsDigit()) {
            unsigned newValue = accum * 10 + peekDigit();
            if (newValue < accum) { /* Overflow check. */
                m_err = QuantifierTooLarge;
                return false;
            }
            accum = newValue;
            consume();
        }
        return true;
    }

    unsigned consumeOctal()
    {
        ASSERT(WTF::isASCIIOctalDigit(peek()));

        unsigned n = consumeDigit();
        while (n < 32 && !atEndOfPattern() && WTF::isASCIIOctalDigit(peek()))
            n = n * 8 + consumeDigit();
        return n;
    }

    bool tryConsume(UChar ch)
    {
        if (atEndOfPattern() || (m_data[m_index] != ch))
            return false;
        ++m_index;
        return true;
    }

    int tryConsumeHex(int count)
    {
        ParseState state = saveState();

        int n = 0;
        while (count--) {
            if (atEndOfPattern() || !WTF::isASCIIHexDigit(peek())) {
                restoreState(state);
                return -1;
            }
            n = (n << 4) | WTF::toASCIIHexValue(consume());
        }
        return n;
    }

    Delegate& m_delegate;
    unsigned m_backReferenceLimit;
    ErrorCode m_err;
    const CharType* m_data;
    unsigned m_size;
    unsigned m_index;
    unsigned m_parenthesesNestingDepth;

    // Derived by empirical testing of compile time in PCRE and WREC.
    static const unsigned MAX_PATTERN_SIZE = 1024 * 1024;
};

/*
 * Yarr::parse():
 *
 * The parse method is passed a pattern to be parsed and a delegate upon which
 * callbacks will be made to record the parsed tokens forming the regex.
 * Yarr::parse() returns null on success, or a const C string providing an error
 * message where a parse error occurs.
 *
 * The Delegate must implement the following interface:
 *
 *    void assertionBOL();
 *    void assertionEOL();
 *    void assertionWordBoundary(bool invert);
 *
 *    void atomPatternCharacter(UChar ch);
 *    void atomBuiltInCharacterClass(BuiltInCharacterClassID classID, bool invert);
 *    void atomCharacterClassBegin(bool invert)
 *    void atomCharacterClassAtom(UChar ch)
 *    void atomCharacterClassRange(UChar begin, UChar end)
 *    void atomCharacterClassBuiltIn(BuiltInCharacterClassID classID, bool invert)
 *    void atomCharacterClassEnd()
 *    void atomParenthesesSubpatternBegin(bool capture = true);
 *    void atomParentheticalAssertionBegin(bool invert = false);
 *    void atomParenthesesEnd();
 *    void atomBackReference(unsigned subpatternId);
 *
 *    void quantifyAtom(unsigned min, unsigned max, bool greedy);
 *
 *    void disjunction();
 *
 * The regular expression is described by a sequence of assertion*() and atom*()
 * callbacks to the delegate, describing the terms in the regular expression.
 * Following an atom a quantifyAtom() call may occur to indicate that the previous
 * atom should be quantified.  In the case of atoms described across multiple
 * calls (parentheses and character classes) the call to quantifyAtom() will come
 * after the call to the atom*End() method, never after atom*Begin().
 *
 * Character classes may either be described by a single call to
 * atomBuiltInCharacterClass(), or by a sequence of atomCharacterClass*() calls.
 * In the latter case, ...Begin() will be called, followed by a sequence of
 * calls to ...Atom(), ...Range(), and ...BuiltIn(), followed by a call to ...End().
 *
 * Sequences of atoms and assertions are broken into alternatives via calls to
 * disjunction().  Assertions, atoms, and disjunctions emitted between calls to
 * atomParenthesesBegin() and atomParenthesesEnd() form the body of a subpattern.
 * atomParenthesesBegin() is passed a subpatternId.  In the case of a regular
 * capturing subpattern, this will be the subpatternId associated with these
 * parentheses, and will also by definition be the lowest subpatternId of these
 * parentheses and of any nested paretheses.  The atomParenthesesEnd() method
 * is passed the subpatternId of the last capturing subexpression nested within
 * these paretheses.  In the case of a capturing subpattern with no nested
 * capturing subpatterns, the same subpatternId will be passed to the begin and
 * end functions.  In the case of non-capturing subpatterns the subpatternId
 * passed to the begin method is also the first possible subpatternId that might
 * be nested within these paretheses.  If a set of non-capturing parentheses does
 * not contain any capturing subpatterns, then the subpatternId passed to begin
 * will be greater than the subpatternId passed to end.
 */

template<class Delegate>
ErrorCode parse(Delegate& delegate, const String& pattern, unsigned backReferenceLimit = quantifyInfinite)
{
#ifdef YARR_8BIT_CHAR_SUPPORT
    if (pattern.is8Bit())
        return Parser<Delegate, LChar>(delegate, pattern, backReferenceLimit).parse();
#endif
    return Parser<Delegate, UChar>(delegate, pattern, backReferenceLimit).parse();
}

} } // namespace JSC::Yarr

#endif /* yarr_YarrParser_h */