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
/* This Source Code Form is subject to the terms of the Mozilla Public
 * License, v. 2.0. If a copy of the MPL was not distributed with this
 * file, You can obtain one at http://mozilla.org/MPL/2.0/. */

use cow_rc_str::CowRcStr;
use smallvec::SmallVec;
use std::ops::BitOr;
use std::ops::Range;
use tokenizer::{SourceLocation, SourcePosition, Token, Tokenizer};

/// A capture of the internal state of a `Parser` (including the position within the input),
/// obtained from the `Parser::position` method.
///
/// Can be used with the `Parser::reset` method to restore that state.
/// Should only be used with the `Parser` instance it came from.
#[derive(Debug, Clone)]
pub struct ParserState {
    pub(crate) position: usize,
    pub(crate) current_line_start_position: usize,
    pub(crate) current_line_number: u32,
    pub(crate) at_start_of: Option<BlockType>,
}

impl ParserState {
    /// The position from the start of the input, counted in UTF-8 bytes.
    #[inline]
    pub fn position(&self) -> SourcePosition {
        SourcePosition(self.position)
    }

    /// The line number and column number
    #[inline]
    pub fn source_location(&self) -> SourceLocation {
        SourceLocation {
            line: self.current_line_number,
            column: (self.position - self.current_line_start_position + 1) as u32,
        }
    }
}

/// Details about a `BasicParseError`
#[derive(Clone, Debug, PartialEq)]
pub enum BasicParseErrorKind<'i> {
    /// An unexpected token was encountered.
    UnexpectedToken(Token<'i>),
    /// The end of the input was encountered unexpectedly.
    EndOfInput,
    /// An `@` rule was encountered that was invalid.
    AtRuleInvalid(CowRcStr<'i>),
    /// The body of an '@' rule was invalid.
    AtRuleBodyInvalid,
    /// A qualified rule was encountered that was invalid.
    QualifiedRuleInvalid,
}

/// The funamental parsing errors that can be triggered by built-in parsing routines.
#[derive(Clone, Debug, PartialEq)]
pub struct BasicParseError<'i> {
    /// Details of this error
    pub kind: BasicParseErrorKind<'i>,
    /// Location where this error occurred
    pub location: SourceLocation,
}

impl<'i, T> From<BasicParseError<'i>> for ParseError<'i, T> {
    #[inline]
    fn from(this: BasicParseError<'i>) -> ParseError<'i, T> {
        ParseError {
            kind: ParseErrorKind::Basic(this.kind),
            location: this.location,
        }
    }
}

impl SourceLocation {
    /// Create a new BasicParseError at this location for an unexpected token
    #[inline]
    pub fn new_basic_unexpected_token_error<'i>(self, token: Token<'i>) -> BasicParseError<'i> {
        BasicParseError {
            kind: BasicParseErrorKind::UnexpectedToken(token),
            location: self,
        }
    }

    /// Create a new ParseError at this location for an unexpected token
    #[inline]
    pub fn new_unexpected_token_error<'i, E>(self, token: Token<'i>) -> ParseError<'i, E> {
        ParseError {
            kind: ParseErrorKind::Basic(BasicParseErrorKind::UnexpectedToken(token)),
            location: self,
        }
    }

    /// Create a new custom ParseError at this location
    #[inline]
    pub fn new_custom_error<'i, E1: Into<E2>, E2>(self, error: E1) -> ParseError<'i, E2> {
        ParseError {
            kind: ParseErrorKind::Custom(error.into()),
            location: self,
        }
    }
}

/// Details of a `ParseError`
#[derive(Clone, Debug, PartialEq)]
pub enum ParseErrorKind<'i, T: 'i> {
    /// A fundamental parse error from a built-in parsing routine.
    Basic(BasicParseErrorKind<'i>),
    /// A parse error reported by downstream consumer code.
    Custom(T),
}

impl<'i, T> ParseErrorKind<'i, T> {
    /// Like `std::convert::Into::into`
    pub fn into<U>(self) -> ParseErrorKind<'i, U>
    where
        T: Into<U>,
    {
        match self {
            ParseErrorKind::Basic(basic) => ParseErrorKind::Basic(basic),
            ParseErrorKind::Custom(custom) => ParseErrorKind::Custom(custom.into()),
        }
    }
}

/// Extensible parse errors that can be encountered by client parsing implementations.
#[derive(Clone, Debug, PartialEq)]
pub struct ParseError<'i, E: 'i> {
    /// Details of this error
    pub kind: ParseErrorKind<'i, E>,
    /// Location where this error occurred
    pub location: SourceLocation,
}

impl<'i, T> ParseError<'i, T> {
    /// Extract the fundamental parse error from an extensible error.
    pub fn basic(self) -> BasicParseError<'i> {
        match self.kind {
            ParseErrorKind::Basic(kind) => BasicParseError {
                kind: kind,
                location: self.location,
            },
            ParseErrorKind::Custom(_) => panic!("Not a basic parse error"),
        }
    }

    /// Like `std::convert::Into::into`
    pub fn into<U>(self) -> ParseError<'i, U>
    where
        T: Into<U>,
    {
        ParseError {
            kind: self.kind.into(),
            location: self.location,
        }
    }
}

/// The owned input for a parser.
pub struct ParserInput<'i> {
    tokenizer: Tokenizer<'i>,
    cached_token: Option<CachedToken<'i>>,
}

struct CachedToken<'i> {
    token: Token<'i>,
    start_position: SourcePosition,
    end_state: ParserState,
}

impl<'i> ParserInput<'i> {
    /// Create a new input for a parser.
    pub fn new(input: &'i str) -> ParserInput<'i> {
        ParserInput {
            tokenizer: Tokenizer::new(input),
            cached_token: None,
        }
    }

    /// Create a new input for a parser.  Line numbers in locations
    /// are offset by the given value.
    pub fn new_with_line_number_offset(input: &'i str, first_line_number: u32) -> ParserInput<'i> {
        ParserInput {
            tokenizer: Tokenizer::with_first_line_number(input, first_line_number),
            cached_token: None,
        }
    }

    #[inline]
    fn cached_token_ref(&self) -> &Token<'i> {
        &self.cached_token.as_ref().unwrap().token
    }
}

/// A CSS parser that borrows its `&str` input,
/// yields `Token`s,
/// and keeps track of nested blocks and functions.
pub struct Parser<'i: 't, 't> {
    input: &'t mut ParserInput<'i>,
    /// If `Some(_)`, .parse_nested_block() can be called.
    at_start_of: Option<BlockType>,
    /// For parsers from `parse_until` or `parse_nested_block`
    stop_before: Delimiters,
}

#[derive(Copy, Clone, PartialEq, Eq, Debug)]
pub(crate) enum BlockType {
    Parenthesis,
    SquareBracket,
    CurlyBracket,
}

impl BlockType {
    fn opening(token: &Token) -> Option<BlockType> {
        match *token {
            Token::Function(_) | Token::ParenthesisBlock => Some(BlockType::Parenthesis),
            Token::SquareBracketBlock => Some(BlockType::SquareBracket),
            Token::CurlyBracketBlock => Some(BlockType::CurlyBracket),
            _ => None,
        }
    }

    fn closing(token: &Token) -> Option<BlockType> {
        match *token {
            Token::CloseParenthesis => Some(BlockType::Parenthesis),
            Token::CloseSquareBracket => Some(BlockType::SquareBracket),
            Token::CloseCurlyBracket => Some(BlockType::CurlyBracket),
            _ => None,
        }
    }
}

/// A set of characters, to be used with the `Parser::parse_until*` methods.
///
/// The union of two sets can be obtained with the `|` operator. Example:
///
/// ```{rust,ignore}
/// input.parse_until_before(Delimiter::CurlyBracketBlock | Delimiter::Semicolon)
/// ```
#[derive(Copy, Clone, PartialEq, Eq, Debug)]
pub struct Delimiters {
    bits: u8,
}

/// `Delimiters` constants.
#[allow(non_upper_case_globals, non_snake_case)]
pub mod Delimiter {
    use super::Delimiters;

    /// The empty delimiter set
    pub const None: Delimiters = Delimiters { bits: 0 };
    /// The delimiter set with only the `{` opening curly bracket
    pub const CurlyBracketBlock: Delimiters = Delimiters { bits: 1 << 1 };
    /// The delimiter set with only the `;` semicolon
    pub const Semicolon: Delimiters = Delimiters { bits: 1 << 2 };
    /// The delimiter set with only the `!` exclamation point
    pub const Bang: Delimiters = Delimiters { bits: 1 << 3 };
    /// The delimiter set with only the `,` comma
    pub const Comma: Delimiters = Delimiters { bits: 1 << 4 };
}

#[allow(non_upper_case_globals, non_snake_case)]
mod ClosingDelimiter {
    use super::Delimiters;

    pub const CloseCurlyBracket: Delimiters = Delimiters { bits: 1 << 5 };
    pub const CloseSquareBracket: Delimiters = Delimiters { bits: 1 << 6 };
    pub const CloseParenthesis: Delimiters = Delimiters { bits: 1 << 7 };
}

impl BitOr<Delimiters> for Delimiters {
    type Output = Delimiters;

    #[inline]
    fn bitor(self, other: Delimiters) -> Delimiters {
        Delimiters {
            bits: self.bits | other.bits,
        }
    }
}

impl Delimiters {
    #[inline]
    fn contains(self, other: Delimiters) -> bool {
        (self.bits & other.bits) != 0
    }

    #[inline]
    fn from_byte(byte: Option<u8>) -> Delimiters {
        match byte {
            Some(b';') => Delimiter::Semicolon,
            Some(b'!') => Delimiter::Bang,
            Some(b',') => Delimiter::Comma,
            Some(b'{') => Delimiter::CurlyBracketBlock,
            Some(b'}') => ClosingDelimiter::CloseCurlyBracket,
            Some(b']') => ClosingDelimiter::CloseSquareBracket,
            Some(b')') => ClosingDelimiter::CloseParenthesis,
            _ => Delimiter::None,
        }
    }
}

/// Used in some `fn expect_*` methods
macro_rules! expect {
    ($parser: ident, $($branches: tt)+) => {
        {
            let start_location = $parser.current_source_location();
            match *$parser.next()? {
                $($branches)+
                ref token => {
                    return Err(start_location.new_basic_unexpected_token_error(token.clone()))
                }
            }
        }
    }
}

impl<'i: 't, 't> Parser<'i, 't> {
    /// Create a new parser
    #[inline]
    pub fn new(input: &'t mut ParserInput<'i>) -> Parser<'i, 't> {
        Parser {
            input: input,
            at_start_of: None,
            stop_before: Delimiter::None,
        }
    }

    /// Return the current line that is being parsed.
    pub fn current_line(&self) -> &'i str {
        self.input.tokenizer.current_source_line()
    }

    /// Check whether the input is exhausted. That is, if `.next()` would return a token.
    ///
    /// This ignores whitespace and comments.
    #[inline]
    pub fn is_exhausted(&mut self) -> bool {
        self.expect_exhausted().is_ok()
    }

    /// Check whether the input is exhausted. That is, if `.next()` would return a token.
    /// Return a `Result` so that the `?` operator can be used: `input.expect_exhausted()?`
    ///
    /// This ignores whitespace and comments.
    #[inline]
    pub fn expect_exhausted(&mut self) -> Result<(), BasicParseError<'i>> {
        let start = self.state();
        let result = match self.next() {
            Err(BasicParseError {
                kind: BasicParseErrorKind::EndOfInput,
                ..
            }) => Ok(()),
            Err(e) => unreachable!("Unexpected error encountered: {:?}", e),
            Ok(t) => Err(start
                .source_location()
                .new_basic_unexpected_token_error(t.clone())),
        };
        self.reset(&start);
        result
    }

    /// Return the current position within the input.
    ///
    /// This can be used with the `Parser::slice` and `slice_from` methods.
    #[inline]
    pub fn position(&self) -> SourcePosition {
        self.input.tokenizer.position()
    }

    /// The current line number and column number.
    #[inline]
    pub fn current_source_location(&self) -> SourceLocation {
        self.input.tokenizer.current_source_location()
    }

    /// The source map URL, if known.
    ///
    /// The source map URL is extracted from a specially formatted
    /// comment.  The last such comment is used, so this value may
    /// change as parsing proceeds.
    pub fn current_source_map_url(&self) -> Option<&str> {
        self.input.tokenizer.current_source_map_url()
    }

    /// The source URL, if known.
    ///
    /// The source URL is extracted from a specially formatted
    /// comment.  The last such comment is used, so this value may
    /// change as parsing proceeds.
    pub fn current_source_url(&self) -> Option<&str> {
        self.input.tokenizer.current_source_url()
    }

    /// Create a new BasicParseError at the current location
    #[inline]
    pub fn new_basic_error(&self, kind: BasicParseErrorKind<'i>) -> BasicParseError<'i> {
        BasicParseError {
            kind: kind,
            location: self.current_source_location(),
        }
    }

    /// Create a new basic ParseError at the current location
    #[inline]
    pub fn new_error<E>(&self, kind: BasicParseErrorKind<'i>) -> ParseError<'i, E> {
        ParseError {
            kind: ParseErrorKind::Basic(kind),
            location: self.current_source_location(),
        }
    }

    /// Create a new custom BasicParseError at the current location
    #[inline]
    pub fn new_custom_error<E1: Into<E2>, E2>(&self, error: E1) -> ParseError<'i, E2> {
        self.current_source_location().new_custom_error(error)
    }

    /// Create a new unexpected token BasicParseError at the current location
    #[inline]
    pub fn new_basic_unexpected_token_error(&self, token: Token<'i>) -> BasicParseError<'i> {
        self.new_basic_error(BasicParseErrorKind::UnexpectedToken(token))
    }

    /// Create a new unexpected token ParseError at the current location
    #[inline]
    pub fn new_unexpected_token_error<E>(&self, token: Token<'i>) -> ParseError<'i, E> {
        self.new_error(BasicParseErrorKind::UnexpectedToken(token))
    }

    /// Create a new unexpected token or EOF ParseError at the current location
    #[inline]
    pub fn new_error_for_next_token<E>(&mut self) -> ParseError<'i, E> {
        let token = match self.next() {
            Ok(token) => token.clone(),
            Err(e) => return e.into(),
        };
        self.new_error(BasicParseErrorKind::UnexpectedToken(token))
    }

    /// Return the current internal state of the parser (including position within the input).
    ///
    /// This state can later be restored with the `Parser::reset` method.
    #[inline]
    pub fn state(&self) -> ParserState {
        ParserState {
            at_start_of: self.at_start_of,
            ..self.input.tokenizer.state()
        }
    }

    /// Advance the input until the next token that’s not whitespace or a comment.
    #[inline]
    pub fn skip_whitespace(&mut self) {
        if let Some(block_type) = self.at_start_of.take() {
            consume_until_end_of_block(block_type, &mut self.input.tokenizer);
        }

        self.input.tokenizer.skip_whitespace()
    }

    #[inline]
    pub(crate) fn skip_cdc_and_cdo(&mut self) {
        if let Some(block_type) = self.at_start_of.take() {
            consume_until_end_of_block(block_type, &mut self.input.tokenizer);
        }

        self.input.tokenizer.skip_cdc_and_cdo()
    }

    #[inline]
    pub(crate) fn next_byte(&self) -> Option<u8> {
        let byte = self.input.tokenizer.next_byte();
        if self.stop_before.contains(Delimiters::from_byte(byte)) {
            return None;
        }
        byte
    }

    /// Restore the internal state of the parser (including position within the input)
    /// to what was previously saved by the `Parser::position` method.
    ///
    /// Should only be used with `SourcePosition` values from the same `Parser` instance.
    #[inline]
    pub fn reset(&mut self, state: &ParserState) {
        self.input.tokenizer.reset(state);
        self.at_start_of = state.at_start_of;
    }

    /// Start looking for `var()` / `env()` functions. (See the
    /// `.seen_var_or_env_functions()` method.)
    #[inline]
    pub fn look_for_var_or_env_functions(&mut self) {
        self.input.tokenizer.look_for_var_or_env_functions()
    }

    /// Return whether a `var()` or `env()` function has been seen by the
    /// tokenizer since either `look_for_var_or_env_functions` was called, and
    /// stop looking.
    #[inline]
    pub fn seen_var_or_env_functions(&mut self) -> bool {
        self.input.tokenizer.seen_var_or_env_functions()
    }

    /// The old name of `try_parse`, which requires raw identifiers in the Rust 2018 edition.
    #[inline]
    pub fn try<F, T, E>(&mut self, thing: F) -> Result<T, E>
    where
        F: FnOnce(&mut Parser<'i, 't>) -> Result<T, E>,
    {
        self.try_parse(thing)
    }

    /// Execute the given closure, passing it the parser.
    /// If the result (returned unchanged) is `Err`,
    /// the internal state of the parser  (including position within the input)
    /// is restored to what it was before the call.
    #[inline]
    pub fn try_parse<F, T, E>(&mut self, thing: F) -> Result<T, E>
    where
        F: FnOnce(&mut Parser<'i, 't>) -> Result<T, E>,
    {
        let start = self.state();
        let result = thing(self);
        if result.is_err() {
            self.reset(&start)
        }
        result
    }

    /// Return a slice of the CSS input
    #[inline]
    pub fn slice(&self, range: Range<SourcePosition>) -> &'i str {
        self.input.tokenizer.slice(range)
    }

    /// Return a slice of the CSS input, from the given position to the current one.
    #[inline]
    pub fn slice_from(&self, start_position: SourcePosition) -> &'i str {
        self.input.tokenizer.slice_from(start_position)
    }

    /// Return the next token in the input that is neither whitespace or a comment,
    /// and advance the position accordingly.
    ///
    /// After returning a `Function`, `ParenthesisBlock`,
    /// `CurlyBracketBlock`, or `SquareBracketBlock` token,
    /// the next call will skip until after the matching `CloseParenthesis`,
    /// `CloseCurlyBracket`, or `CloseSquareBracket` token.
    ///
    /// See the `Parser::parse_nested_block` method to parse the content of functions or blocks.
    ///
    /// This only returns a closing token when it is unmatched (and therefore an error).
    pub fn next(&mut self) -> Result<&Token<'i>, BasicParseError<'i>> {
        self.skip_whitespace();
        self.next_including_whitespace_and_comments()
    }

    /// Same as `Parser::next`, but does not skip whitespace tokens.
    pub fn next_including_whitespace(&mut self) -> Result<&Token<'i>, BasicParseError<'i>> {
        loop {
            match self.next_including_whitespace_and_comments() {
                Err(e) => return Err(e),
                Ok(&Token::Comment(_)) => {}
                _ => break,
            }
        }
        Ok(self.input.cached_token_ref())
    }

    /// Same as `Parser::next`, but does not skip whitespace or comment tokens.
    ///
    /// **Note**: This should only be used in contexts like a CSS pre-processor
    /// where comments are preserved.
    /// When parsing higher-level values, per the CSS Syntax specification,
    /// comments should always be ignored between tokens.
    pub fn next_including_whitespace_and_comments(
        &mut self,
    ) -> Result<&Token<'i>, BasicParseError<'i>> {
        if let Some(block_type) = self.at_start_of.take() {
            consume_until_end_of_block(block_type, &mut self.input.tokenizer);
        }

        let byte = self.input.tokenizer.next_byte();
        if self.stop_before.contains(Delimiters::from_byte(byte)) {
            return Err(self.new_basic_error(BasicParseErrorKind::EndOfInput));
        }

        let token_start_position = self.input.tokenizer.position();
        let using_cached_token = self
            .input
            .cached_token
            .as_ref()
            .map_or(false, |cached_token| {
                cached_token.start_position == token_start_position
            });
        let token = if using_cached_token {
            let cached_token = self.input.cached_token.as_ref().unwrap();
            self.input.tokenizer.reset(&cached_token.end_state);
            match cached_token.token {
                Token::Function(ref name) => self.input.tokenizer.see_function(name),
                _ => {}
            }
            &cached_token.token
        } else {
            let new_token = self
                .input
                .tokenizer
                .next()
                .map_err(|()| self.new_basic_error(BasicParseErrorKind::EndOfInput))?;
            self.input.cached_token = Some(CachedToken {
                token: new_token,
                start_position: token_start_position,
                end_state: self.input.tokenizer.state(),
            });
            self.input.cached_token_ref()
        };

        if let Some(block_type) = BlockType::opening(token) {
            self.at_start_of = Some(block_type);
        }
        Ok(token)
    }

    /// Have the given closure parse something, then check the the input is exhausted.
    /// The result is overridden to `Err(())` if some input remains.
    ///
    /// This can help tell e.g. `color: green;` from `color: green 4px;`
    #[inline]
    pub fn parse_entirely<F, T, E>(&mut self, parse: F) -> Result<T, ParseError<'i, E>>
    where
        F: FnOnce(&mut Parser<'i, 't>) -> Result<T, ParseError<'i, E>>,
    {
        let result = parse(self)?;
        self.expect_exhausted()?;
        Ok(result)
    }

    /// Parse a list of comma-separated values, all with the same syntax.
    ///
    /// The given closure is called repeatedly with a "delimited" parser
    /// (see the `Parser::parse_until_before` method)
    /// so that it can over consume the input past a comma at this block/function nesting level.
    ///
    /// Successful results are accumulated in a vector.
    ///
    /// This method retuns `Err(())` the first time that a closure call does,
    /// or if a closure call leaves some input before the next comma or the end of the input.
    #[inline]
    pub fn parse_comma_separated<F, T, E>(
        &mut self,
        mut parse_one: F,
    ) -> Result<Vec<T>, ParseError<'i, E>>
    where
        F: for<'tt> FnMut(&mut Parser<'i, 'tt>) -> Result<T, ParseError<'i, E>>,
    {
        // Vec grows from 0 to 4 by default on first push().  So allocate with
        // capacity 1, so in the somewhat common case of only one item we don't
        // way overallocate.  Note that we always push at least one item if
        // parsing succeeds.
        let mut values = Vec::with_capacity(1);
        loop {
            self.skip_whitespace(); // Unnecessary for correctness, but may help try() in parse_one rewind less.
            values.push(self.parse_until_before(Delimiter::Comma, &mut parse_one)?);
            match self.next() {
                Err(_) => return Ok(values),
                Ok(&Token::Comma) => continue,
                Ok(_) => unreachable!(),
            }
        }
    }

    /// Parse the content of a block or function.
    ///
    /// This method panics if the last token yielded by this parser
    /// (from one of the `next*` methods)
    /// is not a on that marks the start of a block or function:
    /// a `Function`, `ParenthesisBlock`, `CurlyBracketBlock`, or `SquareBracketBlock`.
    ///
    /// The given closure is called with a "delimited" parser
    /// that stops at the end of the block or function (at the matching closing token).
    ///
    /// The result is overridden to `Err(())` if the closure leaves some input before that point.
    #[inline]
    pub fn parse_nested_block<F, T, E>(&mut self, parse: F) -> Result<T, ParseError<'i, E>>
    where
        F: for<'tt> FnOnce(&mut Parser<'i, 'tt>) -> Result<T, ParseError<'i, E>>,
    {
        parse_nested_block(self, parse)
    }

    /// Limit parsing to until a given delimiter or the end of the input. (E.g.
    /// a semicolon for a property value.)
    ///
    /// The given closure is called with a "delimited" parser
    /// that stops before the first character at this block/function nesting level
    /// that matches the given set of delimiters, or at the end of the input.
    ///
    /// The result is overridden to `Err(())` if the closure leaves some input before that point.
    #[inline]
    pub fn parse_until_before<F, T, E>(
        &mut self,
        delimiters: Delimiters,
        parse: F,
    ) -> Result<T, ParseError<'i, E>>
    where
        F: for<'tt> FnOnce(&mut Parser<'i, 'tt>) -> Result<T, ParseError<'i, E>>,
    {
        parse_until_before(self, delimiters, parse)
    }

    /// Like `parse_until_before`, but also consume the delimiter token.
    ///
    /// This can be useful when you don’t need to know which delimiter it was
    /// (e.g. if these is only one in the given set)
    /// or if it was there at all (as opposed to reaching the end of the input).
    #[inline]
    pub fn parse_until_after<F, T, E>(
        &mut self,
        delimiters: Delimiters,
        parse: F,
    ) -> Result<T, ParseError<'i, E>>
    where
        F: for<'tt> FnOnce(&mut Parser<'i, 'tt>) -> Result<T, ParseError<'i, E>>,
    {
        parse_until_after(self, delimiters, parse)
    }

    /// Parse a <whitespace-token> and return its value.
    #[inline]
    pub fn expect_whitespace(&mut self) -> Result<&'i str, BasicParseError<'i>> {
        let start_location = self.current_source_location();
        match *self.next_including_whitespace()? {
            Token::WhiteSpace(value) => Ok(value),
            ref t => Err(start_location.new_basic_unexpected_token_error(t.clone())),
        }
    }

    /// Parse a <ident-token> and return the unescaped value.
    #[inline]
    pub fn expect_ident(&mut self) -> Result<&CowRcStr<'i>, BasicParseError<'i>> {
        expect! {self,
            Token::Ident(ref value) => Ok(value),
        }
    }

    /// expect_ident, but clone the CowRcStr
    #[inline]
    pub fn expect_ident_cloned(&mut self) -> Result<CowRcStr<'i>, BasicParseError<'i>> {
        self.expect_ident().map(|s| s.clone())
    }

    /// Parse a <ident-token> whose unescaped value is an ASCII-insensitive match for the given value.
    #[inline]
    pub fn expect_ident_matching(
        &mut self,
        expected_value: &str,
    ) -> Result<(), BasicParseError<'i>> {
        expect! {self,
            Token::Ident(ref value) if value.eq_ignore_ascii_case(expected_value) => Ok(()),
        }
    }

    /// Parse a <string-token> and return the unescaped value.
    #[inline]
    pub fn expect_string(&mut self) -> Result<&CowRcStr<'i>, BasicParseError<'i>> {
        expect! {self,
            Token::QuotedString(ref value) => Ok(value),
        }
    }

    /// expect_string, but clone the CowRcStr
    #[inline]
    pub fn expect_string_cloned(&mut self) -> Result<CowRcStr<'i>, BasicParseError<'i>> {
        self.expect_string().map(|s| s.clone())
    }

    /// Parse either a <ident-token> or a <string-token>, and return the unescaped value.
    #[inline]
    pub fn expect_ident_or_string(&mut self) -> Result<&CowRcStr<'i>, BasicParseError<'i>> {
        expect! {self,
            Token::Ident(ref value) => Ok(value),
            Token::QuotedString(ref value) => Ok(value),
        }
    }

    /// Parse a <url-token> and return the unescaped value.
    #[inline]
    pub fn expect_url(&mut self) -> Result<CowRcStr<'i>, BasicParseError<'i>> {
        // FIXME: revert early returns when lifetimes are non-lexical
        expect! {self,
            Token::UnquotedUrl(ref value) => return Ok(value.clone()),
            Token::Function(ref name) if name.eq_ignore_ascii_case("url") => {}
        }
        self.parse_nested_block(|input| {
            input.expect_string().map_err(Into::into).map(|s| s.clone())
        })
        .map_err(ParseError::<()>::basic)
    }

    /// Parse either a <url-token> or a <string-token>, and return the unescaped value.
    #[inline]
    pub fn expect_url_or_string(&mut self) -> Result<CowRcStr<'i>, BasicParseError<'i>> {
        // FIXME: revert early returns when lifetimes are non-lexical
        expect! {self,
            Token::UnquotedUrl(ref value) => return Ok(value.clone()),
            Token::QuotedString(ref value) => return Ok(value.clone()),
            Token::Function(ref name) if name.eq_ignore_ascii_case("url") => {}
        }
        self.parse_nested_block(|input| {
            input.expect_string().map_err(Into::into).map(|s| s.clone())
        })
        .map_err(ParseError::<()>::basic)
    }

    /// Parse a <number-token> and return the integer value.
    #[inline]
    pub fn expect_number(&mut self) -> Result<f32, BasicParseError<'i>> {
        expect! {self,
            Token::Number { value, .. } => Ok(value),
        }
    }

    /// Parse a <number-token> that does not have a fractional part, and return the integer value.
    #[inline]
    pub fn expect_integer(&mut self) -> Result<i32, BasicParseError<'i>> {
        expect! {self,
            Token::Number { int_value: Some(int_value), .. } => Ok(int_value),
        }
    }

    /// Parse a <percentage-token> and return the value.
    /// `0%` and `100%` map to `0.0` and `1.0` (not `100.0`), respectively.
    #[inline]
    pub fn expect_percentage(&mut self) -> Result<f32, BasicParseError<'i>> {
        expect! {self,
            Token::Percentage { unit_value, .. } => Ok(unit_value),
        }
    }

    /// Parse a `:` <colon-token>.
    #[inline]
    pub fn expect_colon(&mut self) -> Result<(), BasicParseError<'i>> {
        expect! {self,
            Token::Colon => Ok(()),
        }
    }

    /// Parse a `;` <semicolon-token>.
    #[inline]
    pub fn expect_semicolon(&mut self) -> Result<(), BasicParseError<'i>> {
        expect! {self,
            Token::Semicolon => Ok(()),
        }
    }

    /// Parse a `,` <comma-token>.
    #[inline]
    pub fn expect_comma(&mut self) -> Result<(), BasicParseError<'i>> {
        expect! {self,
            Token::Comma => Ok(()),
        }
    }

    /// Parse a <delim-token> with the given value.
    #[inline]
    pub fn expect_delim(&mut self, expected_value: char) -> Result<(), BasicParseError<'i>> {
        expect! {self,
            Token::Delim(value) if value == expected_value => Ok(()),
        }
    }

    /// Parse a `{ /* ... */ }` curly brackets block.
    ///
    /// If the result is `Ok`, you can then call the `Parser::parse_nested_block` method.
    #[inline]
    pub fn expect_curly_bracket_block(&mut self) -> Result<(), BasicParseError<'i>> {
        expect! {self,
            Token::CurlyBracketBlock => Ok(()),
        }
    }

    /// Parse a `[ /* ... */ ]` square brackets block.
    ///
    /// If the result is `Ok`, you can then call the `Parser::parse_nested_block` method.
    #[inline]
    pub fn expect_square_bracket_block(&mut self) -> Result<(), BasicParseError<'i>> {
        expect! {self,
            Token::SquareBracketBlock => Ok(()),
        }
    }

    /// Parse a `( /* ... */ )` parenthesis block.
    ///
    /// If the result is `Ok`, you can then call the `Parser::parse_nested_block` method.
    #[inline]
    pub fn expect_parenthesis_block(&mut self) -> Result<(), BasicParseError<'i>> {
        expect! {self,
            Token::ParenthesisBlock => Ok(()),
        }
    }

    /// Parse a <function> token and return its name.
    ///
    /// If the result is `Ok`, you can then call the `Parser::parse_nested_block` method.
    #[inline]
    pub fn expect_function(&mut self) -> Result<&CowRcStr<'i>, BasicParseError<'i>> {
        expect! {self,
            Token::Function(ref name) => Ok(name),
        }
    }

    /// Parse a <function> token whose name is an ASCII-insensitive match for the given value.
    ///
    /// If the result is `Ok`, you can then call the `Parser::parse_nested_block` method.
    #[inline]
    pub fn expect_function_matching(
        &mut self,
        expected_name: &str,
    ) -> Result<(), BasicParseError<'i>> {
        expect! {self,
            Token::Function(ref name) if name.eq_ignore_ascii_case(expected_name) => Ok(()),
        }
    }

    /// Parse the input until exhaustion and check that it contains no “error” token.
    ///
    /// See `Token::is_parse_error`. This also checks nested blocks and functions recursively.
    #[inline]
    pub fn expect_no_error_token(&mut self) -> Result<(), BasicParseError<'i>> {
        // FIXME: remove break and intermediate variable when lifetimes are non-lexical
        let token;
        loop {
            match self.next_including_whitespace_and_comments() {
                Ok(&Token::Function(_))
                | Ok(&Token::ParenthesisBlock)
                | Ok(&Token::SquareBracketBlock)
                | Ok(&Token::CurlyBracketBlock) => {}
                Ok(t) => {
                    if t.is_parse_error() {
                        token = t.clone();
                        break;
                    }
                    continue;
                }
                Err(_) => return Ok(()),
            }
            let result = self.parse_nested_block(|input| {
                input.expect_no_error_token().map_err(|e| Into::into(e))
            });
            result.map_err(ParseError::<()>::basic)?
        }
        // FIXME: maybe these should be separate variants of BasicParseError instead?
        Err(self.new_basic_unexpected_token_error(token))
    }
}

pub fn parse_until_before<'i: 't, 't, F, T, E>(
    parser: &mut Parser<'i, 't>,
    delimiters: Delimiters,
    parse: F,
) -> Result<T, ParseError<'i, E>>
where
    F: for<'tt> FnOnce(&mut Parser<'i, 'tt>) -> Result<T, ParseError<'i, E>>,
{
    let delimiters = parser.stop_before | delimiters;
    let result;
    // Introduce a new scope to limit duration of nested_parser’s borrow
    {
        let mut delimited_parser = Parser {
            input: parser.input,
            at_start_of: parser.at_start_of.take(),
            stop_before: delimiters,
        };
        result = delimited_parser.parse_entirely(parse);
        if let Some(block_type) = delimited_parser.at_start_of {
            consume_until_end_of_block(block_type, &mut delimited_parser.input.tokenizer);
        }
    }
    // FIXME: have a special-purpose tokenizer method for this that does less work.
    loop {
        if delimiters.contains(Delimiters::from_byte(parser.input.tokenizer.next_byte())) {
            break;
        }
        if let Ok(token) = parser.input.tokenizer.next() {
            if let Some(block_type) = BlockType::opening(&token) {
                consume_until_end_of_block(block_type, &mut parser.input.tokenizer);
            }
        } else {
            break;
        }
    }
    result
}

pub fn parse_until_after<'i: 't, 't, F, T, E>(
    parser: &mut Parser<'i, 't>,
    delimiters: Delimiters,
    parse: F,
) -> Result<T, ParseError<'i, E>>
where
    F: for<'tt> FnOnce(&mut Parser<'i, 'tt>) -> Result<T, ParseError<'i, E>>,
{
    let result = parser.parse_until_before(delimiters, parse);
    let next_byte = parser.input.tokenizer.next_byte();
    if next_byte.is_some()
        && !parser
            .stop_before
            .contains(Delimiters::from_byte(next_byte))
    {
        debug_assert!(delimiters.contains(Delimiters::from_byte(next_byte)));
        // We know this byte is ASCII.
        parser.input.tokenizer.advance(1);
        if next_byte == Some(b'{') {
            consume_until_end_of_block(BlockType::CurlyBracket, &mut parser.input.tokenizer);
        }
    }
    result
}

pub fn parse_nested_block<'i: 't, 't, F, T, E>(
    parser: &mut Parser<'i, 't>,
    parse: F,
) -> Result<T, ParseError<'i, E>>
where
    F: for<'tt> FnOnce(&mut Parser<'i, 'tt>) -> Result<T, ParseError<'i, E>>,
{
    let block_type = parser.at_start_of.take().expect(
        "\
         A nested parser can only be created when a Function, \
         ParenthesisBlock, SquareBracketBlock, or CurlyBracketBlock \
         token was just consumed.\
         ",
    );
    let closing_delimiter = match block_type {
        BlockType::CurlyBracket => ClosingDelimiter::CloseCurlyBracket,
        BlockType::SquareBracket => ClosingDelimiter::CloseSquareBracket,
        BlockType::Parenthesis => ClosingDelimiter::CloseParenthesis,
    };
    let result;
    // Introduce a new scope to limit duration of nested_parser’s borrow
    {
        let mut nested_parser = Parser {
            input: parser.input,
            at_start_of: None,
            stop_before: closing_delimiter,
        };
        result = nested_parser.parse_entirely(parse);
        if let Some(block_type) = nested_parser.at_start_of {
            consume_until_end_of_block(block_type, &mut nested_parser.input.tokenizer);
        }
    }
    consume_until_end_of_block(block_type, &mut parser.input.tokenizer);
    result
}

#[inline(never)]
#[cold]
fn consume_until_end_of_block(block_type: BlockType, tokenizer: &mut Tokenizer) {
    let mut stack = SmallVec::<[BlockType; 16]>::new();
    stack.push(block_type);

    // FIXME: have a special-purpose tokenizer method for this that does less work.
    while let Ok(ref token) = tokenizer.next() {
        if let Some(b) = BlockType::closing(token) {
            if *stack.last().unwrap() == b {
                stack.pop();
                if stack.is_empty() {
                    return;
                }
            }
        }

        if let Some(block_type) = BlockType::opening(token) {
            stack.push(block_type);
        }
    }
}