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