sqlglot.tokens

   1from __future__ import annotations
   2
   3import os
   4import typing as t
   5from enum import auto
   6
   7from sqlglot.errors import SqlglotError, TokenError
   8from sqlglot.helper import AutoName
   9from sqlglot.trie import TrieResult, in_trie, new_trie
  10
  11if t.TYPE_CHECKING:
  12    from sqlglot.dialects.dialect import DialectType
  13
  14
  15try:
  16    from sqlglotrs import (  # type: ignore
  17        Tokenizer as RsTokenizer,
  18        TokenizerDialectSettings as RsTokenizerDialectSettings,
  19        TokenizerSettings as RsTokenizerSettings,
  20        TokenTypeSettings as RsTokenTypeSettings,
  21    )
  22
  23    USE_RS_TOKENIZER = os.environ.get("SQLGLOTRS_TOKENIZER", "1") == "1"
  24except ImportError:
  25    USE_RS_TOKENIZER = False
  26
  27
  28class TokenType(AutoName):
  29    L_PAREN = auto()
  30    R_PAREN = auto()
  31    L_BRACKET = auto()
  32    R_BRACKET = auto()
  33    L_BRACE = auto()
  34    R_BRACE = auto()
  35    COMMA = auto()
  36    DOT = auto()
  37    DASH = auto()
  38    PLUS = auto()
  39    COLON = auto()
  40    DCOLON = auto()
  41    DQMARK = auto()
  42    SEMICOLON = auto()
  43    STAR = auto()
  44    BACKSLASH = auto()
  45    SLASH = auto()
  46    LT = auto()
  47    LTE = auto()
  48    GT = auto()
  49    GTE = auto()
  50    NOT = auto()
  51    EQ = auto()
  52    NEQ = auto()
  53    NULLSAFE_EQ = auto()
  54    COLON_EQ = auto()
  55    AND = auto()
  56    OR = auto()
  57    AMP = auto()
  58    DPIPE = auto()
  59    PIPE = auto()
  60    PIPE_SLASH = auto()
  61    DPIPE_SLASH = auto()
  62    CARET = auto()
  63    TILDA = auto()
  64    ARROW = auto()
  65    DARROW = auto()
  66    FARROW = auto()
  67    HASH = auto()
  68    HASH_ARROW = auto()
  69    DHASH_ARROW = auto()
  70    LR_ARROW = auto()
  71    DAT = auto()
  72    LT_AT = auto()
  73    AT_GT = auto()
  74    DOLLAR = auto()
  75    PARAMETER = auto()
  76    SESSION_PARAMETER = auto()
  77    DAMP = auto()
  78    XOR = auto()
  79    DSTAR = auto()
  80
  81    BLOCK_START = auto()
  82    BLOCK_END = auto()
  83
  84    SPACE = auto()
  85    BREAK = auto()
  86
  87    STRING = auto()
  88    NUMBER = auto()
  89    IDENTIFIER = auto()
  90    DATABASE = auto()
  91    COLUMN = auto()
  92    COLUMN_DEF = auto()
  93    SCHEMA = auto()
  94    TABLE = auto()
  95    WAREHOUSE = auto()
  96    STREAMLIT = auto()
  97    VAR = auto()
  98    BIT_STRING = auto()
  99    HEX_STRING = auto()
 100    BYTE_STRING = auto()
 101    NATIONAL_STRING = auto()
 102    RAW_STRING = auto()
 103    HEREDOC_STRING = auto()
 104    UNICODE_STRING = auto()
 105
 106    # types
 107    BIT = auto()
 108    BOOLEAN = auto()
 109    TINYINT = auto()
 110    UTINYINT = auto()
 111    SMALLINT = auto()
 112    USMALLINT = auto()
 113    MEDIUMINT = auto()
 114    UMEDIUMINT = auto()
 115    INT = auto()
 116    UINT = auto()
 117    BIGINT = auto()
 118    UBIGINT = auto()
 119    INT128 = auto()
 120    UINT128 = auto()
 121    INT256 = auto()
 122    UINT256 = auto()
 123    FLOAT = auto()
 124    DOUBLE = auto()
 125    DECIMAL = auto()
 126    DECIMAL32 = auto()
 127    DECIMAL64 = auto()
 128    DECIMAL128 = auto()
 129    UDECIMAL = auto()
 130    BIGDECIMAL = auto()
 131    CHAR = auto()
 132    NCHAR = auto()
 133    VARCHAR = auto()
 134    NVARCHAR = auto()
 135    BPCHAR = auto()
 136    TEXT = auto()
 137    MEDIUMTEXT = auto()
 138    LONGTEXT = auto()
 139    MEDIUMBLOB = auto()
 140    LONGBLOB = auto()
 141    TINYBLOB = auto()
 142    TINYTEXT = auto()
 143    NAME = auto()
 144    BINARY = auto()
 145    VARBINARY = auto()
 146    JSON = auto()
 147    JSONB = auto()
 148    TIME = auto()
 149    TIMETZ = auto()
 150    TIMESTAMP = auto()
 151    TIMESTAMPTZ = auto()
 152    TIMESTAMPLTZ = auto()
 153    TIMESTAMPNTZ = auto()
 154    TIMESTAMP_S = auto()
 155    TIMESTAMP_MS = auto()
 156    TIMESTAMP_NS = auto()
 157    DATETIME = auto()
 158    DATETIME64 = auto()
 159    DATE = auto()
 160    DATE32 = auto()
 161    INT4RANGE = auto()
 162    INT4MULTIRANGE = auto()
 163    INT8RANGE = auto()
 164    INT8MULTIRANGE = auto()
 165    NUMRANGE = auto()
 166    NUMMULTIRANGE = auto()
 167    TSRANGE = auto()
 168    TSMULTIRANGE = auto()
 169    TSTZRANGE = auto()
 170    TSTZMULTIRANGE = auto()
 171    DATERANGE = auto()
 172    DATEMULTIRANGE = auto()
 173    UUID = auto()
 174    GEOGRAPHY = auto()
 175    NULLABLE = auto()
 176    GEOMETRY = auto()
 177    HLLSKETCH = auto()
 178    HSTORE = auto()
 179    SUPER = auto()
 180    SERIAL = auto()
 181    SMALLSERIAL = auto()
 182    BIGSERIAL = auto()
 183    XML = auto()
 184    YEAR = auto()
 185    UNIQUEIDENTIFIER = auto()
 186    USERDEFINED = auto()
 187    MONEY = auto()
 188    SMALLMONEY = auto()
 189    ROWVERSION = auto()
 190    IMAGE = auto()
 191    VARIANT = auto()
 192    OBJECT = auto()
 193    INET = auto()
 194    IPADDRESS = auto()
 195    IPPREFIX = auto()
 196    IPV4 = auto()
 197    IPV6 = auto()
 198    ENUM = auto()
 199    ENUM8 = auto()
 200    ENUM16 = auto()
 201    FIXEDSTRING = auto()
 202    LOWCARDINALITY = auto()
 203    NESTED = auto()
 204    AGGREGATEFUNCTION = auto()
 205    SIMPLEAGGREGATEFUNCTION = auto()
 206    TDIGEST = auto()
 207    UNKNOWN = auto()
 208    VECTOR = auto()
 209
 210    # keywords
 211    ALIAS = auto()
 212    ALTER = auto()
 213    ALWAYS = auto()
 214    ALL = auto()
 215    ANTI = auto()
 216    ANY = auto()
 217    APPLY = auto()
 218    ARRAY = auto()
 219    ASC = auto()
 220    ASOF = auto()
 221    AUTO_INCREMENT = auto()
 222    BEGIN = auto()
 223    BETWEEN = auto()
 224    CACHE = auto()
 225    CASE = auto()
 226    CHARACTER_SET = auto()
 227    CLUSTER_BY = auto()
 228    COLLATE = auto()
 229    COMMAND = auto()
 230    COMMENT = auto()
 231    COMMIT = auto()
 232    CONNECT_BY = auto()
 233    CONSTRAINT = auto()
 234    COPY = auto()
 235    CREATE = auto()
 236    CROSS = auto()
 237    CUBE = auto()
 238    CURRENT_DATE = auto()
 239    CURRENT_DATETIME = auto()
 240    CURRENT_TIME = auto()
 241    CURRENT_TIMESTAMP = auto()
 242    CURRENT_USER = auto()
 243    DECLARE = auto()
 244    DEFAULT = auto()
 245    DELETE = auto()
 246    DESC = auto()
 247    DESCRIBE = auto()
 248    DICTIONARY = auto()
 249    DISTINCT = auto()
 250    DISTRIBUTE_BY = auto()
 251    DIV = auto()
 252    DROP = auto()
 253    ELSE = auto()
 254    END = auto()
 255    ESCAPE = auto()
 256    EXCEPT = auto()
 257    EXECUTE = auto()
 258    EXISTS = auto()
 259    FALSE = auto()
 260    FETCH = auto()
 261    FILTER = auto()
 262    FINAL = auto()
 263    FIRST = auto()
 264    FOR = auto()
 265    FORCE = auto()
 266    FOREIGN_KEY = auto()
 267    FORMAT = auto()
 268    FROM = auto()
 269    FULL = auto()
 270    FUNCTION = auto()
 271    GLOB = auto()
 272    GLOBAL = auto()
 273    GROUP_BY = auto()
 274    GROUPING_SETS = auto()
 275    HAVING = auto()
 276    HINT = auto()
 277    IGNORE = auto()
 278    ILIKE = auto()
 279    ILIKE_ANY = auto()
 280    IN = auto()
 281    INDEX = auto()
 282    INNER = auto()
 283    INSERT = auto()
 284    INTERSECT = auto()
 285    INTERVAL = auto()
 286    INTO = auto()
 287    INTRODUCER = auto()
 288    IRLIKE = auto()
 289    IS = auto()
 290    ISNULL = auto()
 291    JOIN = auto()
 292    JOIN_MARKER = auto()
 293    KEEP = auto()
 294    KEY = auto()
 295    KILL = auto()
 296    LANGUAGE = auto()
 297    LATERAL = auto()
 298    LEFT = auto()
 299    LIKE = auto()
 300    LIKE_ANY = auto()
 301    LIMIT = auto()
 302    LIST = auto()
 303    LOAD = auto()
 304    LOCK = auto()
 305    MAP = auto()
 306    MATCH_CONDITION = auto()
 307    MATCH_RECOGNIZE = auto()
 308    MEMBER_OF = auto()
 309    MERGE = auto()
 310    MOD = auto()
 311    MODEL = auto()
 312    NATURAL = auto()
 313    NEXT = auto()
 314    NOTNULL = auto()
 315    NULL = auto()
 316    OBJECT_IDENTIFIER = auto()
 317    OFFSET = auto()
 318    ON = auto()
 319    ONLY = auto()
 320    OPERATOR = auto()
 321    ORDER_BY = auto()
 322    ORDER_SIBLINGS_BY = auto()
 323    ORDERED = auto()
 324    ORDINALITY = auto()
 325    OUTER = auto()
 326    OVER = auto()
 327    OVERLAPS = auto()
 328    OVERWRITE = auto()
 329    PARTITION = auto()
 330    PARTITION_BY = auto()
 331    PERCENT = auto()
 332    PIVOT = auto()
 333    PLACEHOLDER = auto()
 334    POSITIONAL = auto()
 335    PRAGMA = auto()
 336    PREWHERE = auto()
 337    PRIMARY_KEY = auto()
 338    PROCEDURE = auto()
 339    PROPERTIES = auto()
 340    PSEUDO_TYPE = auto()
 341    QUALIFY = auto()
 342    QUOTE = auto()
 343    RANGE = auto()
 344    RECURSIVE = auto()
 345    REFRESH = auto()
 346    RENAME = auto()
 347    REPLACE = auto()
 348    RETURNING = auto()
 349    REFERENCES = auto()
 350    RIGHT = auto()
 351    RLIKE = auto()
 352    ROLLBACK = auto()
 353    ROLLUP = auto()
 354    ROW = auto()
 355    ROWS = auto()
 356    SELECT = auto()
 357    SEMI = auto()
 358    SEPARATOR = auto()
 359    SEQUENCE = auto()
 360    SERDE_PROPERTIES = auto()
 361    SET = auto()
 362    SETTINGS = auto()
 363    SHOW = auto()
 364    SIMILAR_TO = auto()
 365    SOME = auto()
 366    SORT_BY = auto()
 367    START_WITH = auto()
 368    STORAGE_INTEGRATION = auto()
 369    STRAIGHT_JOIN = auto()
 370    STRUCT = auto()
 371    SUMMARIZE = auto()
 372    TABLE_SAMPLE = auto()
 373    TAG = auto()
 374    TEMPORARY = auto()
 375    TOP = auto()
 376    THEN = auto()
 377    TRUE = auto()
 378    TRUNCATE = auto()
 379    UNCACHE = auto()
 380    UNION = auto()
 381    UNNEST = auto()
 382    UNPIVOT = auto()
 383    UPDATE = auto()
 384    USE = auto()
 385    USING = auto()
 386    VALUES = auto()
 387    VIEW = auto()
 388    VOLATILE = auto()
 389    WHEN = auto()
 390    WHERE = auto()
 391    WINDOW = auto()
 392    WITH = auto()
 393    UNIQUE = auto()
 394    VERSION_SNAPSHOT = auto()
 395    TIMESTAMP_SNAPSHOT = auto()
 396    OPTION = auto()
 397
 398
 399_ALL_TOKEN_TYPES = list(TokenType)
 400_TOKEN_TYPE_TO_INDEX = {token_type: i for i, token_type in enumerate(_ALL_TOKEN_TYPES)}
 401
 402
 403class Token:
 404    __slots__ = ("token_type", "text", "line", "col", "start", "end", "comments")
 405
 406    @classmethod
 407    def number(cls, number: int) -> Token:
 408        """Returns a NUMBER token with `number` as its text."""
 409        return cls(TokenType.NUMBER, str(number))
 410
 411    @classmethod
 412    def string(cls, string: str) -> Token:
 413        """Returns a STRING token with `string` as its text."""
 414        return cls(TokenType.STRING, string)
 415
 416    @classmethod
 417    def identifier(cls, identifier: str) -> Token:
 418        """Returns an IDENTIFIER token with `identifier` as its text."""
 419        return cls(TokenType.IDENTIFIER, identifier)
 420
 421    @classmethod
 422    def var(cls, var: str) -> Token:
 423        """Returns an VAR token with `var` as its text."""
 424        return cls(TokenType.VAR, var)
 425
 426    def __init__(
 427        self,
 428        token_type: TokenType,
 429        text: str,
 430        line: int = 1,
 431        col: int = 1,
 432        start: int = 0,
 433        end: int = 0,
 434        comments: t.Optional[t.List[str]] = None,
 435    ) -> None:
 436        """Token initializer.
 437
 438        Args:
 439            token_type: The TokenType Enum.
 440            text: The text of the token.
 441            line: The line that the token ends on.
 442            col: The column that the token ends on.
 443            start: The start index of the token.
 444            end: The ending index of the token.
 445            comments: The comments to attach to the token.
 446        """
 447        self.token_type = token_type
 448        self.text = text
 449        self.line = line
 450        self.col = col
 451        self.start = start
 452        self.end = end
 453        self.comments = [] if comments is None else comments
 454
 455    def __repr__(self) -> str:
 456        attributes = ", ".join(f"{k}: {getattr(self, k)}" for k in self.__slots__)
 457        return f"<Token {attributes}>"
 458
 459
 460class _Tokenizer(type):
 461    def __new__(cls, clsname, bases, attrs):
 462        klass = super().__new__(cls, clsname, bases, attrs)
 463
 464        def _convert_quotes(arr: t.List[str | t.Tuple[str, str]]) -> t.Dict[str, str]:
 465            return dict(
 466                (item, item) if isinstance(item, str) else (item[0], item[1]) for item in arr
 467            )
 468
 469        def _quotes_to_format(
 470            token_type: TokenType, arr: t.List[str | t.Tuple[str, str]]
 471        ) -> t.Dict[str, t.Tuple[str, TokenType]]:
 472            return {k: (v, token_type) for k, v in _convert_quotes(arr).items()}
 473
 474        klass._QUOTES = _convert_quotes(klass.QUOTES)
 475        klass._IDENTIFIERS = _convert_quotes(klass.IDENTIFIERS)
 476
 477        klass._FORMAT_STRINGS = {
 478            **{
 479                p + s: (e, TokenType.NATIONAL_STRING)
 480                for s, e in klass._QUOTES.items()
 481                for p in ("n", "N")
 482            },
 483            **_quotes_to_format(TokenType.BIT_STRING, klass.BIT_STRINGS),
 484            **_quotes_to_format(TokenType.BYTE_STRING, klass.BYTE_STRINGS),
 485            **_quotes_to_format(TokenType.HEX_STRING, klass.HEX_STRINGS),
 486            **_quotes_to_format(TokenType.RAW_STRING, klass.RAW_STRINGS),
 487            **_quotes_to_format(TokenType.HEREDOC_STRING, klass.HEREDOC_STRINGS),
 488            **_quotes_to_format(TokenType.UNICODE_STRING, klass.UNICODE_STRINGS),
 489        }
 490
 491        klass._STRING_ESCAPES = set(klass.STRING_ESCAPES)
 492        klass._IDENTIFIER_ESCAPES = set(klass.IDENTIFIER_ESCAPES)
 493        klass._COMMENTS = {
 494            **dict(
 495                (comment, None) if isinstance(comment, str) else (comment[0], comment[1])
 496                for comment in klass.COMMENTS
 497            ),
 498            "{#": "#}",  # Ensure Jinja comments are tokenized correctly in all dialects
 499        }
 500
 501        klass._KEYWORD_TRIE = new_trie(
 502            key.upper()
 503            for key in (
 504                *klass.KEYWORDS,
 505                *klass._COMMENTS,
 506                *klass._QUOTES,
 507                *klass._FORMAT_STRINGS,
 508            )
 509            if " " in key or any(single in key for single in klass.SINGLE_TOKENS)
 510        )
 511
 512        if USE_RS_TOKENIZER:
 513            settings = RsTokenizerSettings(
 514                white_space={k: _TOKEN_TYPE_TO_INDEX[v] for k, v in klass.WHITE_SPACE.items()},
 515                single_tokens={k: _TOKEN_TYPE_TO_INDEX[v] for k, v in klass.SINGLE_TOKENS.items()},
 516                keywords={k: _TOKEN_TYPE_TO_INDEX[v] for k, v in klass.KEYWORDS.items()},
 517                numeric_literals=klass.NUMERIC_LITERALS,
 518                identifiers=klass._IDENTIFIERS,
 519                identifier_escapes=klass._IDENTIFIER_ESCAPES,
 520                string_escapes=klass._STRING_ESCAPES,
 521                quotes=klass._QUOTES,
 522                format_strings={
 523                    k: (v1, _TOKEN_TYPE_TO_INDEX[v2])
 524                    for k, (v1, v2) in klass._FORMAT_STRINGS.items()
 525                },
 526                has_bit_strings=bool(klass.BIT_STRINGS),
 527                has_hex_strings=bool(klass.HEX_STRINGS),
 528                comments=klass._COMMENTS,
 529                var_single_tokens=klass.VAR_SINGLE_TOKENS,
 530                commands={_TOKEN_TYPE_TO_INDEX[v] for v in klass.COMMANDS},
 531                command_prefix_tokens={
 532                    _TOKEN_TYPE_TO_INDEX[v] for v in klass.COMMAND_PREFIX_TOKENS
 533                },
 534                heredoc_tag_is_identifier=klass.HEREDOC_TAG_IS_IDENTIFIER,
 535                string_escapes_allowed_in_raw_strings=klass.STRING_ESCAPES_ALLOWED_IN_RAW_STRINGS,
 536                nested_comments=klass.NESTED_COMMENTS,
 537            )
 538            token_types = RsTokenTypeSettings(
 539                bit_string=_TOKEN_TYPE_TO_INDEX[TokenType.BIT_STRING],
 540                break_=_TOKEN_TYPE_TO_INDEX[TokenType.BREAK],
 541                dcolon=_TOKEN_TYPE_TO_INDEX[TokenType.DCOLON],
 542                heredoc_string=_TOKEN_TYPE_TO_INDEX[TokenType.HEREDOC_STRING],
 543                raw_string=_TOKEN_TYPE_TO_INDEX[TokenType.RAW_STRING],
 544                hex_string=_TOKEN_TYPE_TO_INDEX[TokenType.HEX_STRING],
 545                identifier=_TOKEN_TYPE_TO_INDEX[TokenType.IDENTIFIER],
 546                number=_TOKEN_TYPE_TO_INDEX[TokenType.NUMBER],
 547                parameter=_TOKEN_TYPE_TO_INDEX[TokenType.PARAMETER],
 548                semicolon=_TOKEN_TYPE_TO_INDEX[TokenType.SEMICOLON],
 549                string=_TOKEN_TYPE_TO_INDEX[TokenType.STRING],
 550                var=_TOKEN_TYPE_TO_INDEX[TokenType.VAR],
 551                heredoc_string_alternative=_TOKEN_TYPE_TO_INDEX[klass.HEREDOC_STRING_ALTERNATIVE],
 552            )
 553            klass._RS_TOKENIZER = RsTokenizer(settings, token_types)
 554        else:
 555            klass._RS_TOKENIZER = None
 556
 557        return klass
 558
 559
 560class Tokenizer(metaclass=_Tokenizer):
 561    SINGLE_TOKENS = {
 562        "(": TokenType.L_PAREN,
 563        ")": TokenType.R_PAREN,
 564        "[": TokenType.L_BRACKET,
 565        "]": TokenType.R_BRACKET,
 566        "{": TokenType.L_BRACE,
 567        "}": TokenType.R_BRACE,
 568        "&": TokenType.AMP,
 569        "^": TokenType.CARET,
 570        ":": TokenType.COLON,
 571        ",": TokenType.COMMA,
 572        ".": TokenType.DOT,
 573        "-": TokenType.DASH,
 574        "=": TokenType.EQ,
 575        ">": TokenType.GT,
 576        "<": TokenType.LT,
 577        "%": TokenType.MOD,
 578        "!": TokenType.NOT,
 579        "|": TokenType.PIPE,
 580        "+": TokenType.PLUS,
 581        ";": TokenType.SEMICOLON,
 582        "/": TokenType.SLASH,
 583        "\\": TokenType.BACKSLASH,
 584        "*": TokenType.STAR,
 585        "~": TokenType.TILDA,
 586        "?": TokenType.PLACEHOLDER,
 587        "@": TokenType.PARAMETER,
 588        "#": TokenType.HASH,
 589        # Used for breaking a var like x'y' but nothing else the token type doesn't matter
 590        "'": TokenType.UNKNOWN,
 591        "`": TokenType.UNKNOWN,
 592        '"': TokenType.UNKNOWN,
 593    }
 594
 595    BIT_STRINGS: t.List[str | t.Tuple[str, str]] = []
 596    BYTE_STRINGS: t.List[str | t.Tuple[str, str]] = []
 597    HEX_STRINGS: t.List[str | t.Tuple[str, str]] = []
 598    RAW_STRINGS: t.List[str | t.Tuple[str, str]] = []
 599    HEREDOC_STRINGS: t.List[str | t.Tuple[str, str]] = []
 600    UNICODE_STRINGS: t.List[str | t.Tuple[str, str]] = []
 601    IDENTIFIERS: t.List[str | t.Tuple[str, str]] = ['"']
 602    IDENTIFIER_ESCAPES = ['"']
 603    QUOTES: t.List[t.Tuple[str, str] | str] = ["'"]
 604    STRING_ESCAPES = ["'"]
 605    VAR_SINGLE_TOKENS: t.Set[str] = set()
 606
 607    # Whether the heredoc tags follow the same lexical rules as unquoted identifiers
 608    HEREDOC_TAG_IS_IDENTIFIER = False
 609
 610    # Token that we'll generate as a fallback if the heredoc prefix doesn't correspond to a heredoc
 611    HEREDOC_STRING_ALTERNATIVE = TokenType.VAR
 612
 613    # Whether string escape characters function as such when placed within raw strings
 614    STRING_ESCAPES_ALLOWED_IN_RAW_STRINGS = True
 615
 616    NESTED_COMMENTS = True
 617
 618    # Autofilled
 619    _COMMENTS: t.Dict[str, str] = {}
 620    _FORMAT_STRINGS: t.Dict[str, t.Tuple[str, TokenType]] = {}
 621    _IDENTIFIERS: t.Dict[str, str] = {}
 622    _IDENTIFIER_ESCAPES: t.Set[str] = set()
 623    _QUOTES: t.Dict[str, str] = {}
 624    _STRING_ESCAPES: t.Set[str] = set()
 625    _KEYWORD_TRIE: t.Dict = {}
 626    _RS_TOKENIZER: t.Optional[t.Any] = None
 627
 628    KEYWORDS: t.Dict[str, TokenType] = {
 629        **{f"{{%{postfix}": TokenType.BLOCK_START for postfix in ("", "+", "-")},
 630        **{f"{prefix}%}}": TokenType.BLOCK_END for prefix in ("", "+", "-")},
 631        **{f"{{{{{postfix}": TokenType.BLOCK_START for postfix in ("+", "-")},
 632        **{f"{prefix}}}}}": TokenType.BLOCK_END for prefix in ("+", "-")},
 633        "/*+": TokenType.HINT,
 634        "==": TokenType.EQ,
 635        "::": TokenType.DCOLON,
 636        "||": TokenType.DPIPE,
 637        ">=": TokenType.GTE,
 638        "<=": TokenType.LTE,
 639        "<>": TokenType.NEQ,
 640        "!=": TokenType.NEQ,
 641        ":=": TokenType.COLON_EQ,
 642        "<=>": TokenType.NULLSAFE_EQ,
 643        "->": TokenType.ARROW,
 644        "->>": TokenType.DARROW,
 645        "=>": TokenType.FARROW,
 646        "#>": TokenType.HASH_ARROW,
 647        "#>>": TokenType.DHASH_ARROW,
 648        "<->": TokenType.LR_ARROW,
 649        "&&": TokenType.DAMP,
 650        "??": TokenType.DQMARK,
 651        "ALL": TokenType.ALL,
 652        "ALWAYS": TokenType.ALWAYS,
 653        "AND": TokenType.AND,
 654        "ANTI": TokenType.ANTI,
 655        "ANY": TokenType.ANY,
 656        "ASC": TokenType.ASC,
 657        "AS": TokenType.ALIAS,
 658        "ASOF": TokenType.ASOF,
 659        "AUTOINCREMENT": TokenType.AUTO_INCREMENT,
 660        "AUTO_INCREMENT": TokenType.AUTO_INCREMENT,
 661        "BEGIN": TokenType.BEGIN,
 662        "BETWEEN": TokenType.BETWEEN,
 663        "CACHE": TokenType.CACHE,
 664        "UNCACHE": TokenType.UNCACHE,
 665        "CASE": TokenType.CASE,
 666        "CHARACTER SET": TokenType.CHARACTER_SET,
 667        "CLUSTER BY": TokenType.CLUSTER_BY,
 668        "COLLATE": TokenType.COLLATE,
 669        "COLUMN": TokenType.COLUMN,
 670        "COMMIT": TokenType.COMMIT,
 671        "CONNECT BY": TokenType.CONNECT_BY,
 672        "CONSTRAINT": TokenType.CONSTRAINT,
 673        "COPY": TokenType.COPY,
 674        "CREATE": TokenType.CREATE,
 675        "CROSS": TokenType.CROSS,
 676        "CUBE": TokenType.CUBE,
 677        "CURRENT_DATE": TokenType.CURRENT_DATE,
 678        "CURRENT_TIME": TokenType.CURRENT_TIME,
 679        "CURRENT_TIMESTAMP": TokenType.CURRENT_TIMESTAMP,
 680        "CURRENT_USER": TokenType.CURRENT_USER,
 681        "DATABASE": TokenType.DATABASE,
 682        "DEFAULT": TokenType.DEFAULT,
 683        "DELETE": TokenType.DELETE,
 684        "DESC": TokenType.DESC,
 685        "DESCRIBE": TokenType.DESCRIBE,
 686        "DISTINCT": TokenType.DISTINCT,
 687        "DISTRIBUTE BY": TokenType.DISTRIBUTE_BY,
 688        "DIV": TokenType.DIV,
 689        "DROP": TokenType.DROP,
 690        "ELSE": TokenType.ELSE,
 691        "END": TokenType.END,
 692        "ENUM": TokenType.ENUM,
 693        "ESCAPE": TokenType.ESCAPE,
 694        "EXCEPT": TokenType.EXCEPT,
 695        "EXECUTE": TokenType.EXECUTE,
 696        "EXISTS": TokenType.EXISTS,
 697        "FALSE": TokenType.FALSE,
 698        "FETCH": TokenType.FETCH,
 699        "FILTER": TokenType.FILTER,
 700        "FIRST": TokenType.FIRST,
 701        "FULL": TokenType.FULL,
 702        "FUNCTION": TokenType.FUNCTION,
 703        "FOR": TokenType.FOR,
 704        "FOREIGN KEY": TokenType.FOREIGN_KEY,
 705        "FORMAT": TokenType.FORMAT,
 706        "FROM": TokenType.FROM,
 707        "GEOGRAPHY": TokenType.GEOGRAPHY,
 708        "GEOMETRY": TokenType.GEOMETRY,
 709        "GLOB": TokenType.GLOB,
 710        "GROUP BY": TokenType.GROUP_BY,
 711        "GROUPING SETS": TokenType.GROUPING_SETS,
 712        "HAVING": TokenType.HAVING,
 713        "ILIKE": TokenType.ILIKE,
 714        "IN": TokenType.IN,
 715        "INDEX": TokenType.INDEX,
 716        "INET": TokenType.INET,
 717        "INNER": TokenType.INNER,
 718        "INSERT": TokenType.INSERT,
 719        "INTERVAL": TokenType.INTERVAL,
 720        "INTERSECT": TokenType.INTERSECT,
 721        "INTO": TokenType.INTO,
 722        "IS": TokenType.IS,
 723        "ISNULL": TokenType.ISNULL,
 724        "JOIN": TokenType.JOIN,
 725        "KEEP": TokenType.KEEP,
 726        "KILL": TokenType.KILL,
 727        "LATERAL": TokenType.LATERAL,
 728        "LEFT": TokenType.LEFT,
 729        "LIKE": TokenType.LIKE,
 730        "LIMIT": TokenType.LIMIT,
 731        "LOAD": TokenType.LOAD,
 732        "LOCK": TokenType.LOCK,
 733        "MERGE": TokenType.MERGE,
 734        "NATURAL": TokenType.NATURAL,
 735        "NEXT": TokenType.NEXT,
 736        "NOT": TokenType.NOT,
 737        "NOTNULL": TokenType.NOTNULL,
 738        "NULL": TokenType.NULL,
 739        "OBJECT": TokenType.OBJECT,
 740        "OFFSET": TokenType.OFFSET,
 741        "ON": TokenType.ON,
 742        "OR": TokenType.OR,
 743        "XOR": TokenType.XOR,
 744        "ORDER BY": TokenType.ORDER_BY,
 745        "ORDINALITY": TokenType.ORDINALITY,
 746        "OUTER": TokenType.OUTER,
 747        "OVER": TokenType.OVER,
 748        "OVERLAPS": TokenType.OVERLAPS,
 749        "OVERWRITE": TokenType.OVERWRITE,
 750        "PARTITION": TokenType.PARTITION,
 751        "PARTITION BY": TokenType.PARTITION_BY,
 752        "PARTITIONED BY": TokenType.PARTITION_BY,
 753        "PARTITIONED_BY": TokenType.PARTITION_BY,
 754        "PERCENT": TokenType.PERCENT,
 755        "PIVOT": TokenType.PIVOT,
 756        "PRAGMA": TokenType.PRAGMA,
 757        "PRIMARY KEY": TokenType.PRIMARY_KEY,
 758        "PROCEDURE": TokenType.PROCEDURE,
 759        "QUALIFY": TokenType.QUALIFY,
 760        "RANGE": TokenType.RANGE,
 761        "RECURSIVE": TokenType.RECURSIVE,
 762        "REGEXP": TokenType.RLIKE,
 763        "RENAME": TokenType.RENAME,
 764        "REPLACE": TokenType.REPLACE,
 765        "RETURNING": TokenType.RETURNING,
 766        "REFERENCES": TokenType.REFERENCES,
 767        "RIGHT": TokenType.RIGHT,
 768        "RLIKE": TokenType.RLIKE,
 769        "ROLLBACK": TokenType.ROLLBACK,
 770        "ROLLUP": TokenType.ROLLUP,
 771        "ROW": TokenType.ROW,
 772        "ROWS": TokenType.ROWS,
 773        "SCHEMA": TokenType.SCHEMA,
 774        "SELECT": TokenType.SELECT,
 775        "SEMI": TokenType.SEMI,
 776        "SET": TokenType.SET,
 777        "SETTINGS": TokenType.SETTINGS,
 778        "SHOW": TokenType.SHOW,
 779        "SIMILAR TO": TokenType.SIMILAR_TO,
 780        "SOME": TokenType.SOME,
 781        "SORT BY": TokenType.SORT_BY,
 782        "START WITH": TokenType.START_WITH,
 783        "STRAIGHT_JOIN": TokenType.STRAIGHT_JOIN,
 784        "TABLE": TokenType.TABLE,
 785        "TABLESAMPLE": TokenType.TABLE_SAMPLE,
 786        "TEMP": TokenType.TEMPORARY,
 787        "TEMPORARY": TokenType.TEMPORARY,
 788        "THEN": TokenType.THEN,
 789        "TRUE": TokenType.TRUE,
 790        "TRUNCATE": TokenType.TRUNCATE,
 791        "UNION": TokenType.UNION,
 792        "UNKNOWN": TokenType.UNKNOWN,
 793        "UNNEST": TokenType.UNNEST,
 794        "UNPIVOT": TokenType.UNPIVOT,
 795        "UPDATE": TokenType.UPDATE,
 796        "USE": TokenType.USE,
 797        "USING": TokenType.USING,
 798        "UUID": TokenType.UUID,
 799        "VALUES": TokenType.VALUES,
 800        "VIEW": TokenType.VIEW,
 801        "VOLATILE": TokenType.VOLATILE,
 802        "WHEN": TokenType.WHEN,
 803        "WHERE": TokenType.WHERE,
 804        "WINDOW": TokenType.WINDOW,
 805        "WITH": TokenType.WITH,
 806        "APPLY": TokenType.APPLY,
 807        "ARRAY": TokenType.ARRAY,
 808        "BIT": TokenType.BIT,
 809        "BOOL": TokenType.BOOLEAN,
 810        "BOOLEAN": TokenType.BOOLEAN,
 811        "BYTE": TokenType.TINYINT,
 812        "MEDIUMINT": TokenType.MEDIUMINT,
 813        "INT1": TokenType.TINYINT,
 814        "TINYINT": TokenType.TINYINT,
 815        "INT16": TokenType.SMALLINT,
 816        "SHORT": TokenType.SMALLINT,
 817        "SMALLINT": TokenType.SMALLINT,
 818        "INT128": TokenType.INT128,
 819        "HUGEINT": TokenType.INT128,
 820        "INT2": TokenType.SMALLINT,
 821        "INTEGER": TokenType.INT,
 822        "INT": TokenType.INT,
 823        "INT4": TokenType.INT,
 824        "INT32": TokenType.INT,
 825        "INT64": TokenType.BIGINT,
 826        "LONG": TokenType.BIGINT,
 827        "BIGINT": TokenType.BIGINT,
 828        "INT8": TokenType.TINYINT,
 829        "UINT": TokenType.UINT,
 830        "DEC": TokenType.DECIMAL,
 831        "DECIMAL": TokenType.DECIMAL,
 832        "DECIMAL32": TokenType.DECIMAL32,
 833        "DECIMAL64": TokenType.DECIMAL64,
 834        "DECIMAL128": TokenType.DECIMAL128,
 835        "BIGDECIMAL": TokenType.BIGDECIMAL,
 836        "BIGNUMERIC": TokenType.BIGDECIMAL,
 837        "LIST": TokenType.LIST,
 838        "MAP": TokenType.MAP,
 839        "NULLABLE": TokenType.NULLABLE,
 840        "NUMBER": TokenType.DECIMAL,
 841        "NUMERIC": TokenType.DECIMAL,
 842        "FIXED": TokenType.DECIMAL,
 843        "REAL": TokenType.FLOAT,
 844        "FLOAT": TokenType.FLOAT,
 845        "FLOAT4": TokenType.FLOAT,
 846        "FLOAT8": TokenType.DOUBLE,
 847        "DOUBLE": TokenType.DOUBLE,
 848        "DOUBLE PRECISION": TokenType.DOUBLE,
 849        "JSON": TokenType.JSON,
 850        "JSONB": TokenType.JSONB,
 851        "CHAR": TokenType.CHAR,
 852        "CHARACTER": TokenType.CHAR,
 853        "NCHAR": TokenType.NCHAR,
 854        "VARCHAR": TokenType.VARCHAR,
 855        "VARCHAR2": TokenType.VARCHAR,
 856        "NVARCHAR": TokenType.NVARCHAR,
 857        "NVARCHAR2": TokenType.NVARCHAR,
 858        "BPCHAR": TokenType.BPCHAR,
 859        "STR": TokenType.TEXT,
 860        "STRING": TokenType.TEXT,
 861        "TEXT": TokenType.TEXT,
 862        "LONGTEXT": TokenType.LONGTEXT,
 863        "MEDIUMTEXT": TokenType.MEDIUMTEXT,
 864        "TINYTEXT": TokenType.TINYTEXT,
 865        "CLOB": TokenType.TEXT,
 866        "LONGVARCHAR": TokenType.TEXT,
 867        "BINARY": TokenType.BINARY,
 868        "BLOB": TokenType.VARBINARY,
 869        "LONGBLOB": TokenType.LONGBLOB,
 870        "MEDIUMBLOB": TokenType.MEDIUMBLOB,
 871        "TINYBLOB": TokenType.TINYBLOB,
 872        "BYTEA": TokenType.VARBINARY,
 873        "VARBINARY": TokenType.VARBINARY,
 874        "TIME": TokenType.TIME,
 875        "TIMETZ": TokenType.TIMETZ,
 876        "TIMESTAMP": TokenType.TIMESTAMP,
 877        "TIMESTAMPTZ": TokenType.TIMESTAMPTZ,
 878        "TIMESTAMPLTZ": TokenType.TIMESTAMPLTZ,
 879        "TIMESTAMP_LTZ": TokenType.TIMESTAMPLTZ,
 880        "TIMESTAMPNTZ": TokenType.TIMESTAMPNTZ,
 881        "TIMESTAMP_NTZ": TokenType.TIMESTAMPNTZ,
 882        "DATE": TokenType.DATE,
 883        "DATETIME": TokenType.DATETIME,
 884        "INT4RANGE": TokenType.INT4RANGE,
 885        "INT4MULTIRANGE": TokenType.INT4MULTIRANGE,
 886        "INT8RANGE": TokenType.INT8RANGE,
 887        "INT8MULTIRANGE": TokenType.INT8MULTIRANGE,
 888        "NUMRANGE": TokenType.NUMRANGE,
 889        "NUMMULTIRANGE": TokenType.NUMMULTIRANGE,
 890        "TSRANGE": TokenType.TSRANGE,
 891        "TSMULTIRANGE": TokenType.TSMULTIRANGE,
 892        "TSTZRANGE": TokenType.TSTZRANGE,
 893        "TSTZMULTIRANGE": TokenType.TSTZMULTIRANGE,
 894        "DATERANGE": TokenType.DATERANGE,
 895        "DATEMULTIRANGE": TokenType.DATEMULTIRANGE,
 896        "UNIQUE": TokenType.UNIQUE,
 897        "VECTOR": TokenType.VECTOR,
 898        "STRUCT": TokenType.STRUCT,
 899        "SEQUENCE": TokenType.SEQUENCE,
 900        "VARIANT": TokenType.VARIANT,
 901        "ALTER": TokenType.ALTER,
 902        "ANALYZE": TokenType.COMMAND,
 903        "CALL": TokenType.COMMAND,
 904        "COMMENT": TokenType.COMMENT,
 905        "EXPLAIN": TokenType.COMMAND,
 906        "GRANT": TokenType.COMMAND,
 907        "OPTIMIZE": TokenType.COMMAND,
 908        "PREPARE": TokenType.COMMAND,
 909        "VACUUM": TokenType.COMMAND,
 910        "USER-DEFINED": TokenType.USERDEFINED,
 911        "FOR VERSION": TokenType.VERSION_SNAPSHOT,
 912        "FOR TIMESTAMP": TokenType.TIMESTAMP_SNAPSHOT,
 913    }
 914
 915    WHITE_SPACE: t.Dict[t.Optional[str], TokenType] = {
 916        " ": TokenType.SPACE,
 917        "\t": TokenType.SPACE,
 918        "\n": TokenType.BREAK,
 919        "\r": TokenType.BREAK,
 920    }
 921
 922    COMMANDS = {
 923        TokenType.COMMAND,
 924        TokenType.EXECUTE,
 925        TokenType.FETCH,
 926        TokenType.SHOW,
 927        TokenType.RENAME,
 928    }
 929
 930    COMMAND_PREFIX_TOKENS = {TokenType.SEMICOLON, TokenType.BEGIN}
 931
 932    # Handle numeric literals like in hive (3L = BIGINT)
 933    NUMERIC_LITERALS: t.Dict[str, str] = {}
 934
 935    COMMENTS = ["--", ("/*", "*/")]
 936
 937    __slots__ = (
 938        "sql",
 939        "size",
 940        "tokens",
 941        "dialect",
 942        "_start",
 943        "_current",
 944        "_line",
 945        "_col",
 946        "_comments",
 947        "_char",
 948        "_end",
 949        "_peek",
 950        "_prev_token_line",
 951        "_rs_dialect_settings",
 952    )
 953
 954    def __init__(self, dialect: DialectType = None) -> None:
 955        from sqlglot.dialects import Dialect
 956
 957        self.dialect = Dialect.get_or_raise(dialect)
 958
 959        if USE_RS_TOKENIZER:
 960            self._rs_dialect_settings = RsTokenizerDialectSettings(
 961                unescaped_sequences=self.dialect.UNESCAPED_SEQUENCES,
 962                identifiers_can_start_with_digit=self.dialect.IDENTIFIERS_CAN_START_WITH_DIGIT,
 963            )
 964
 965        self.reset()
 966
 967    def reset(self) -> None:
 968        self.sql = ""
 969        self.size = 0
 970        self.tokens: t.List[Token] = []
 971        self._start = 0
 972        self._current = 0
 973        self._line = 1
 974        self._col = 0
 975        self._comments: t.List[str] = []
 976
 977        self._char = ""
 978        self._end = False
 979        self._peek = ""
 980        self._prev_token_line = -1
 981
 982    def tokenize(self, sql: str) -> t.List[Token]:
 983        """Returns a list of tokens corresponding to the SQL string `sql`."""
 984        if USE_RS_TOKENIZER:
 985            return self.tokenize_rs(sql)
 986
 987        self.reset()
 988        self.sql = sql
 989        self.size = len(sql)
 990
 991        try:
 992            self._scan()
 993        except Exception as e:
 994            start = max(self._current - 50, 0)
 995            end = min(self._current + 50, self.size - 1)
 996            context = self.sql[start:end]
 997            raise TokenError(f"Error tokenizing '{context}'") from e
 998
 999        return self.tokens
1000
1001    def _scan(self, until: t.Optional[t.Callable] = None) -> None:
1002        while self.size and not self._end:
1003            current = self._current
1004
1005            # Skip spaces here rather than iteratively calling advance() for performance reasons
1006            while current < self.size:
1007                char = self.sql[current]
1008
1009                if char.isspace() and (char == " " or char == "\t"):
1010                    current += 1
1011                else:
1012                    break
1013
1014            offset = current - self._current if current > self._current else 1
1015
1016            self._start = current
1017            self._advance(offset)
1018
1019            if not self._char.isspace():
1020                if self._char.isdigit():
1021                    self._scan_number()
1022                elif self._char in self._IDENTIFIERS:
1023                    self._scan_identifier(self._IDENTIFIERS[self._char])
1024                else:
1025                    self._scan_keywords()
1026
1027            if until and until():
1028                break
1029
1030        if self.tokens and self._comments:
1031            self.tokens[-1].comments.extend(self._comments)
1032
1033    def _chars(self, size: int) -> str:
1034        if size == 1:
1035            return self._char
1036
1037        start = self._current - 1
1038        end = start + size
1039
1040        return self.sql[start:end] if end <= self.size else ""
1041
1042    def _advance(self, i: int = 1, alnum: bool = False) -> None:
1043        if self.WHITE_SPACE.get(self._char) is TokenType.BREAK:
1044            # Ensures we don't count an extra line if we get a \r\n line break sequence
1045            if not (self._char == "\r" and self._peek == "\n"):
1046                self._col = 1
1047                self._line += 1
1048        else:
1049            self._col += i
1050
1051        self._current += i
1052        self._end = self._current >= self.size
1053        self._char = self.sql[self._current - 1]
1054        self._peek = "" if self._end else self.sql[self._current]
1055
1056        if alnum and self._char.isalnum():
1057            # Here we use local variables instead of attributes for better performance
1058            _col = self._col
1059            _current = self._current
1060            _end = self._end
1061            _peek = self._peek
1062
1063            while _peek.isalnum():
1064                _col += 1
1065                _current += 1
1066                _end = _current >= self.size
1067                _peek = "" if _end else self.sql[_current]
1068
1069            self._col = _col
1070            self._current = _current
1071            self._end = _end
1072            self._peek = _peek
1073            self._char = self.sql[_current - 1]
1074
1075    @property
1076    def _text(self) -> str:
1077        return self.sql[self._start : self._current]
1078
1079    def _add(self, token_type: TokenType, text: t.Optional[str] = None) -> None:
1080        self._prev_token_line = self._line
1081
1082        if self._comments and token_type == TokenType.SEMICOLON and self.tokens:
1083            self.tokens[-1].comments.extend(self._comments)
1084            self._comments = []
1085
1086        self.tokens.append(
1087            Token(
1088                token_type,
1089                text=self._text if text is None else text,
1090                line=self._line,
1091                col=self._col,
1092                start=self._start,
1093                end=self._current - 1,
1094                comments=self._comments,
1095            )
1096        )
1097        self._comments = []
1098
1099        # If we have either a semicolon or a begin token before the command's token, we'll parse
1100        # whatever follows the command's token as a string
1101        if (
1102            token_type in self.COMMANDS
1103            and self._peek != ";"
1104            and (len(self.tokens) == 1 or self.tokens[-2].token_type in self.COMMAND_PREFIX_TOKENS)
1105        ):
1106            start = self._current
1107            tokens = len(self.tokens)
1108            self._scan(lambda: self._peek == ";")
1109            self.tokens = self.tokens[:tokens]
1110            text = self.sql[start : self._current].strip()
1111            if text:
1112                self._add(TokenType.STRING, text)
1113
1114    def _scan_keywords(self) -> None:
1115        size = 0
1116        word = None
1117        chars = self._text
1118        char = chars
1119        prev_space = False
1120        skip = False
1121        trie = self._KEYWORD_TRIE
1122        single_token = char in self.SINGLE_TOKENS
1123
1124        while chars:
1125            if skip:
1126                result = TrieResult.PREFIX
1127            else:
1128                result, trie = in_trie(trie, char.upper())
1129
1130            if result == TrieResult.FAILED:
1131                break
1132            if result == TrieResult.EXISTS:
1133                word = chars
1134
1135            end = self._current + size
1136            size += 1
1137
1138            if end < self.size:
1139                char = self.sql[end]
1140                single_token = single_token or char in self.SINGLE_TOKENS
1141                is_space = char.isspace()
1142
1143                if not is_space or not prev_space:
1144                    if is_space:
1145                        char = " "
1146                    chars += char
1147                    prev_space = is_space
1148                    skip = False
1149                else:
1150                    skip = True
1151            else:
1152                char = ""
1153                break
1154
1155        if word:
1156            if self._scan_string(word):
1157                return
1158            if self._scan_comment(word):
1159                return
1160            if prev_space or single_token or not char:
1161                self._advance(size - 1)
1162                word = word.upper()
1163                self._add(self.KEYWORDS[word], text=word)
1164                return
1165
1166        if self._char in self.SINGLE_TOKENS:
1167            self._add(self.SINGLE_TOKENS[self._char], text=self._char)
1168            return
1169
1170        self._scan_var()
1171
1172    def _scan_comment(self, comment_start: str) -> bool:
1173        if comment_start not in self._COMMENTS:
1174            return False
1175
1176        comment_start_line = self._line
1177        comment_start_size = len(comment_start)
1178        comment_end = self._COMMENTS[comment_start]
1179
1180        if comment_end:
1181            # Skip the comment's start delimiter
1182            self._advance(comment_start_size)
1183
1184            comment_count = 1
1185            comment_end_size = len(comment_end)
1186
1187            while not self._end:
1188                if self._chars(comment_end_size) == comment_end:
1189                    comment_count -= 1
1190                    if not comment_count:
1191                        break
1192
1193                self._advance(alnum=True)
1194
1195                # Nested comments are allowed by some dialects, e.g. databricks, duckdb, postgres
1196                if (
1197                    self.NESTED_COMMENTS
1198                    and not self._end
1199                    and self._chars(comment_end_size) == comment_start
1200                ):
1201                    self._advance(comment_start_size)
1202                    comment_count += 1
1203
1204            self._comments.append(self._text[comment_start_size : -comment_end_size + 1])
1205            self._advance(comment_end_size - 1)
1206        else:
1207            while not self._end and self.WHITE_SPACE.get(self._peek) is not TokenType.BREAK:
1208                self._advance(alnum=True)
1209            self._comments.append(self._text[comment_start_size:])
1210
1211        # Leading comment is attached to the succeeding token, whilst trailing comment to the preceding.
1212        # Multiple consecutive comments are preserved by appending them to the current comments list.
1213        if comment_start_line == self._prev_token_line:
1214            self.tokens[-1].comments.extend(self._comments)
1215            self._comments = []
1216            self._prev_token_line = self._line
1217
1218        return True
1219
1220    def _scan_number(self) -> None:
1221        if self._char == "0":
1222            peek = self._peek.upper()
1223            if peek == "B":
1224                return self._scan_bits() if self.BIT_STRINGS else self._add(TokenType.NUMBER)
1225            elif peek == "X":
1226                return self._scan_hex() if self.HEX_STRINGS else self._add(TokenType.NUMBER)
1227
1228        decimal = False
1229        scientific = 0
1230
1231        while True:
1232            if self._peek.isdigit():
1233                self._advance()
1234            elif self._peek == "." and not decimal:
1235                if self.tokens and self.tokens[-1].token_type == TokenType.PARAMETER:
1236                    return self._add(TokenType.NUMBER)
1237                decimal = True
1238                self._advance()
1239            elif self._peek in ("-", "+") and scientific == 1:
1240                scientific += 1
1241                self._advance()
1242            elif self._peek.upper() == "E" and not scientific:
1243                scientific += 1
1244                self._advance()
1245            elif self._peek.isidentifier():
1246                number_text = self._text
1247                literal = ""
1248
1249                while self._peek.strip() and self._peek not in self.SINGLE_TOKENS:
1250                    literal += self._peek
1251                    self._advance()
1252
1253                token_type = self.KEYWORDS.get(self.NUMERIC_LITERALS.get(literal.upper(), ""))
1254
1255                if token_type:
1256                    self._add(TokenType.NUMBER, number_text)
1257                    self._add(TokenType.DCOLON, "::")
1258                    return self._add(token_type, literal)
1259                elif self.dialect.IDENTIFIERS_CAN_START_WITH_DIGIT:
1260                    return self._add(TokenType.VAR)
1261
1262                self._advance(-len(literal))
1263                return self._add(TokenType.NUMBER, number_text)
1264            else:
1265                return self._add(TokenType.NUMBER)
1266
1267    def _scan_bits(self) -> None:
1268        self._advance()
1269        value = self._extract_value()
1270        try:
1271            # If `value` can't be converted to a binary, fallback to tokenizing it as an identifier
1272            int(value, 2)
1273            self._add(TokenType.BIT_STRING, value[2:])  # Drop the 0b
1274        except ValueError:
1275            self._add(TokenType.IDENTIFIER)
1276
1277    def _scan_hex(self) -> None:
1278        self._advance()
1279        value = self._extract_value()
1280        try:
1281            # If `value` can't be converted to a hex, fallback to tokenizing it as an identifier
1282            int(value, 16)
1283            self._add(TokenType.HEX_STRING, value[2:])  # Drop the 0x
1284        except ValueError:
1285            self._add(TokenType.IDENTIFIER)
1286
1287    def _extract_value(self) -> str:
1288        while True:
1289            char = self._peek.strip()
1290            if char and char not in self.SINGLE_TOKENS:
1291                self._advance(alnum=True)
1292            else:
1293                break
1294
1295        return self._text
1296
1297    def _scan_string(self, start: str) -> bool:
1298        base = None
1299        token_type = TokenType.STRING
1300
1301        if start in self._QUOTES:
1302            end = self._QUOTES[start]
1303        elif start in self._FORMAT_STRINGS:
1304            end, token_type = self._FORMAT_STRINGS[start]
1305
1306            if token_type == TokenType.HEX_STRING:
1307                base = 16
1308            elif token_type == TokenType.BIT_STRING:
1309                base = 2
1310            elif token_type == TokenType.HEREDOC_STRING:
1311                self._advance()
1312
1313                if self._char == end:
1314                    tag = ""
1315                else:
1316                    tag = self._extract_string(
1317                        end,
1318                        raw_string=True,
1319                        raise_unmatched=not self.HEREDOC_TAG_IS_IDENTIFIER,
1320                    )
1321
1322                if tag and self.HEREDOC_TAG_IS_IDENTIFIER and (self._end or not tag.isidentifier()):
1323                    if not self._end:
1324                        self._advance(-1)
1325
1326                    self._advance(-len(tag))
1327                    self._add(self.HEREDOC_STRING_ALTERNATIVE)
1328                    return True
1329
1330                end = f"{start}{tag}{end}"
1331        else:
1332            return False
1333
1334        self._advance(len(start))
1335        text = self._extract_string(end, raw_string=token_type == TokenType.RAW_STRING)
1336
1337        if base:
1338            try:
1339                int(text, base)
1340            except Exception:
1341                raise TokenError(
1342                    f"Numeric string contains invalid characters from {self._line}:{self._start}"
1343                )
1344
1345        self._add(token_type, text)
1346        return True
1347
1348    def _scan_identifier(self, identifier_end: str) -> None:
1349        self._advance()
1350        text = self._extract_string(identifier_end, escapes=self._IDENTIFIER_ESCAPES)
1351        self._add(TokenType.IDENTIFIER, text)
1352
1353    def _scan_var(self) -> None:
1354        while True:
1355            char = self._peek.strip()
1356            if char and (char in self.VAR_SINGLE_TOKENS or char not in self.SINGLE_TOKENS):
1357                self._advance(alnum=True)
1358            else:
1359                break
1360
1361        self._add(
1362            TokenType.VAR
1363            if self.tokens and self.tokens[-1].token_type == TokenType.PARAMETER
1364            else self.KEYWORDS.get(self._text.upper(), TokenType.VAR)
1365        )
1366
1367    def _extract_string(
1368        self,
1369        delimiter: str,
1370        escapes: t.Optional[t.Set[str]] = None,
1371        raw_string: bool = False,
1372        raise_unmatched: bool = True,
1373    ) -> str:
1374        text = ""
1375        delim_size = len(delimiter)
1376        escapes = self._STRING_ESCAPES if escapes is None else escapes
1377
1378        while True:
1379            if (
1380                not raw_string
1381                and self.dialect.UNESCAPED_SEQUENCES
1382                and self._peek
1383                and self._char in self.STRING_ESCAPES
1384            ):
1385                unescaped_sequence = self.dialect.UNESCAPED_SEQUENCES.get(self._char + self._peek)
1386                if unescaped_sequence:
1387                    self._advance(2)
1388                    text += unescaped_sequence
1389                    continue
1390            if (
1391                (self.STRING_ESCAPES_ALLOWED_IN_RAW_STRINGS or not raw_string)
1392                and self._char in escapes
1393                and (self._peek == delimiter or self._peek in escapes)
1394                and (self._char not in self._QUOTES or self._char == self._peek)
1395            ):
1396                if self._peek == delimiter:
1397                    text += self._peek
1398                else:
1399                    text += self._char + self._peek
1400
1401                if self._current + 1 < self.size:
1402                    self._advance(2)
1403                else:
1404                    raise TokenError(f"Missing {delimiter} from {self._line}:{self._current}")
1405            else:
1406                if self._chars(delim_size) == delimiter:
1407                    if delim_size > 1:
1408                        self._advance(delim_size - 1)
1409                    break
1410
1411                if self._end:
1412                    if not raise_unmatched:
1413                        return text + self._char
1414
1415                    raise TokenError(f"Missing {delimiter} from {self._line}:{self._start}")
1416
1417                current = self._current - 1
1418                self._advance(alnum=True)
1419                text += self.sql[current : self._current - 1]
1420
1421        return text
1422
1423    def tokenize_rs(self, sql: str) -> t.List[Token]:
1424        if not self._RS_TOKENIZER:
1425            raise SqlglotError("Rust tokenizer is not available")
1426
1427        try:
1428            tokens = self._RS_TOKENIZER.tokenize(sql, self._rs_dialect_settings)
1429            for token in tokens:
1430                token.token_type = _ALL_TOKEN_TYPES[token.token_type_index]
1431            return tokens
1432        except Exception as e:
1433            raise TokenError(str(e))