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    CARET_AT = auto()
  64    TILDA = auto()
  65    ARROW = auto()
  66    DARROW = auto()
  67    FARROW = auto()
  68    HASH = auto()
  69    HASH_ARROW = auto()
  70    DHASH_ARROW = auto()
  71    LR_ARROW = auto()
  72    DAT = auto()
  73    LT_AT = auto()
  74    AT_GT = auto()
  75    DOLLAR = auto()
  76    PARAMETER = auto()
  77    SESSION_PARAMETER = auto()
  78    DAMP = auto()
  79    XOR = auto()
  80    DSTAR = auto()
  81
  82    BLOCK_START = auto()
  83    BLOCK_END = auto()
  84
  85    SPACE = auto()
  86    BREAK = auto()
  87
  88    STRING = auto()
  89    NUMBER = auto()
  90    IDENTIFIER = auto()
  91    DATABASE = auto()
  92    COLUMN = auto()
  93    COLUMN_DEF = auto()
  94    SCHEMA = auto()
  95    TABLE = auto()
  96    WAREHOUSE = auto()
  97    STREAMLIT = auto()
  98    VAR = auto()
  99    BIT_STRING = auto()
 100    HEX_STRING = auto()
 101    BYTE_STRING = auto()
 102    NATIONAL_STRING = auto()
 103    RAW_STRING = auto()
 104    HEREDOC_STRING = auto()
 105    UNICODE_STRING = auto()
 106
 107    # types
 108    BIT = auto()
 109    BOOLEAN = auto()
 110    TINYINT = auto()
 111    UTINYINT = auto()
 112    SMALLINT = auto()
 113    USMALLINT = auto()
 114    MEDIUMINT = auto()
 115    UMEDIUMINT = auto()
 116    INT = auto()
 117    UINT = auto()
 118    BIGINT = auto()
 119    UBIGINT = auto()
 120    INT128 = auto()
 121    UINT128 = auto()
 122    INT256 = auto()
 123    UINT256 = auto()
 124    FLOAT = auto()
 125    DOUBLE = auto()
 126    DECIMAL = auto()
 127    DECIMAL32 = auto()
 128    DECIMAL64 = auto()
 129    DECIMAL128 = auto()
 130    DECIMAL256 = auto()
 131    UDECIMAL = auto()
 132    BIGDECIMAL = auto()
 133    CHAR = auto()
 134    NCHAR = auto()
 135    VARCHAR = auto()
 136    NVARCHAR = auto()
 137    BPCHAR = auto()
 138    TEXT = auto()
 139    MEDIUMTEXT = auto()
 140    LONGTEXT = auto()
 141    MEDIUMBLOB = auto()
 142    LONGBLOB = auto()
 143    TINYBLOB = auto()
 144    TINYTEXT = auto()
 145    NAME = auto()
 146    BINARY = auto()
 147    VARBINARY = auto()
 148    JSON = auto()
 149    JSONB = auto()
 150    TIME = auto()
 151    TIMETZ = auto()
 152    TIMESTAMP = auto()
 153    TIMESTAMPTZ = auto()
 154    TIMESTAMPLTZ = auto()
 155    TIMESTAMPNTZ = auto()
 156    TIMESTAMP_S = auto()
 157    TIMESTAMP_MS = auto()
 158    TIMESTAMP_NS = auto()
 159    DATETIME = auto()
 160    DATETIME64 = auto()
 161    DATE = auto()
 162    DATE32 = auto()
 163    INT4RANGE = auto()
 164    INT4MULTIRANGE = auto()
 165    INT8RANGE = auto()
 166    INT8MULTIRANGE = auto()
 167    NUMRANGE = auto()
 168    NUMMULTIRANGE = auto()
 169    TSRANGE = auto()
 170    TSMULTIRANGE = auto()
 171    TSTZRANGE = auto()
 172    TSTZMULTIRANGE = auto()
 173    DATERANGE = auto()
 174    DATEMULTIRANGE = auto()
 175    UUID = auto()
 176    GEOGRAPHY = auto()
 177    NULLABLE = auto()
 178    GEOMETRY = auto()
 179    POINT = auto()
 180    RING = auto()
 181    LINESTRING = auto()
 182    MULTILINESTRING = auto()
 183    POLYGON = auto()
 184    MULTIPOLYGON = auto()
 185    HLLSKETCH = auto()
 186    HSTORE = auto()
 187    SUPER = auto()
 188    SERIAL = auto()
 189    SMALLSERIAL = auto()
 190    BIGSERIAL = auto()
 191    XML = auto()
 192    YEAR = auto()
 193    UNIQUEIDENTIFIER = auto()
 194    USERDEFINED = auto()
 195    MONEY = auto()
 196    SMALLMONEY = auto()
 197    ROWVERSION = auto()
 198    IMAGE = auto()
 199    VARIANT = auto()
 200    OBJECT = auto()
 201    INET = auto()
 202    IPADDRESS = auto()
 203    IPPREFIX = auto()
 204    IPV4 = auto()
 205    IPV6 = auto()
 206    ENUM = auto()
 207    ENUM8 = auto()
 208    ENUM16 = auto()
 209    FIXEDSTRING = auto()
 210    LOWCARDINALITY = auto()
 211    NESTED = auto()
 212    AGGREGATEFUNCTION = auto()
 213    SIMPLEAGGREGATEFUNCTION = auto()
 214    TDIGEST = auto()
 215    UNKNOWN = auto()
 216    VECTOR = auto()
 217
 218    # keywords
 219    ALIAS = auto()
 220    ALTER = auto()
 221    ALWAYS = auto()
 222    ALL = auto()
 223    ANTI = auto()
 224    ANY = auto()
 225    APPLY = auto()
 226    ARRAY = auto()
 227    ASC = auto()
 228    ASOF = auto()
 229    ATTACH = auto()
 230    AUTO_INCREMENT = auto()
 231    BEGIN = auto()
 232    BETWEEN = auto()
 233    BULK_COLLECT_INTO = auto()
 234    CACHE = auto()
 235    CASE = auto()
 236    CHARACTER_SET = auto()
 237    CLUSTER_BY = auto()
 238    COLLATE = auto()
 239    COMMAND = auto()
 240    COMMENT = auto()
 241    COMMIT = auto()
 242    CONNECT_BY = auto()
 243    CONSTRAINT = auto()
 244    COPY = auto()
 245    CREATE = auto()
 246    CROSS = auto()
 247    CUBE = auto()
 248    CURRENT_DATE = auto()
 249    CURRENT_DATETIME = auto()
 250    CURRENT_TIME = auto()
 251    CURRENT_TIMESTAMP = auto()
 252    CURRENT_USER = auto()
 253    DECLARE = auto()
 254    DEFAULT = auto()
 255    DELETE = auto()
 256    DESC = auto()
 257    DESCRIBE = auto()
 258    DETACH = auto()
 259    DICTIONARY = auto()
 260    DISTINCT = auto()
 261    DISTRIBUTE_BY = auto()
 262    DIV = auto()
 263    DROP = auto()
 264    ELSE = auto()
 265    END = auto()
 266    ESCAPE = auto()
 267    EXCEPT = auto()
 268    EXECUTE = auto()
 269    EXISTS = auto()
 270    FALSE = auto()
 271    FETCH = auto()
 272    FILTER = auto()
 273    FINAL = auto()
 274    FIRST = auto()
 275    FOR = auto()
 276    FORCE = auto()
 277    FOREIGN_KEY = auto()
 278    FORMAT = auto()
 279    FROM = auto()
 280    FULL = auto()
 281    FUNCTION = auto()
 282    GLOB = auto()
 283    GLOBAL = auto()
 284    GRANT = auto()
 285    GROUP_BY = auto()
 286    GROUPING_SETS = auto()
 287    HAVING = auto()
 288    HINT = auto()
 289    IGNORE = auto()
 290    ILIKE = auto()
 291    ILIKE_ANY = auto()
 292    IN = auto()
 293    INDEX = auto()
 294    INNER = auto()
 295    INSERT = auto()
 296    INTERSECT = auto()
 297    INTERVAL = auto()
 298    INTO = auto()
 299    INTRODUCER = auto()
 300    IRLIKE = auto()
 301    IS = auto()
 302    ISNULL = auto()
 303    JOIN = auto()
 304    JOIN_MARKER = auto()
 305    KEEP = auto()
 306    KEY = auto()
 307    KILL = auto()
 308    LANGUAGE = auto()
 309    LATERAL = auto()
 310    LEFT = auto()
 311    LIKE = auto()
 312    LIKE_ANY = auto()
 313    LIMIT = auto()
 314    LIST = auto()
 315    LOAD = auto()
 316    LOCK = auto()
 317    MAP = auto()
 318    MATCH_CONDITION = auto()
 319    MATCH_RECOGNIZE = auto()
 320    MEMBER_OF = auto()
 321    MERGE = auto()
 322    MOD = auto()
 323    MODEL = auto()
 324    NATURAL = auto()
 325    NEXT = auto()
 326    NOTNULL = auto()
 327    NULL = auto()
 328    OBJECT_IDENTIFIER = auto()
 329    OFFSET = auto()
 330    ON = auto()
 331    ONLY = auto()
 332    OPERATOR = auto()
 333    ORDER_BY = auto()
 334    ORDER_SIBLINGS_BY = auto()
 335    ORDERED = auto()
 336    ORDINALITY = auto()
 337    OUTER = auto()
 338    OVER = auto()
 339    OVERLAPS = auto()
 340    OVERWRITE = auto()
 341    PARTITION = auto()
 342    PARTITION_BY = auto()
 343    PERCENT = auto()
 344    PIVOT = auto()
 345    PLACEHOLDER = auto()
 346    POSITIONAL = auto()
 347    PRAGMA = auto()
 348    PREWHERE = auto()
 349    PRIMARY_KEY = auto()
 350    PROCEDURE = auto()
 351    PROPERTIES = auto()
 352    PSEUDO_TYPE = auto()
 353    QUALIFY = auto()
 354    QUOTE = auto()
 355    RANGE = auto()
 356    RECURSIVE = auto()
 357    REFRESH = auto()
 358    RENAME = auto()
 359    REPLACE = auto()
 360    RETURNING = auto()
 361    REFERENCES = auto()
 362    RIGHT = auto()
 363    RLIKE = auto()
 364    ROLLBACK = auto()
 365    ROLLUP = auto()
 366    ROW = auto()
 367    ROWS = auto()
 368    SELECT = auto()
 369    SEMI = auto()
 370    SEPARATOR = auto()
 371    SEQUENCE = auto()
 372    SERDE_PROPERTIES = auto()
 373    SET = auto()
 374    SETTINGS = auto()
 375    SHOW = auto()
 376    SIMILAR_TO = auto()
 377    SOME = auto()
 378    SORT_BY = auto()
 379    START_WITH = auto()
 380    STORAGE_INTEGRATION = auto()
 381    STRAIGHT_JOIN = auto()
 382    STRUCT = auto()
 383    SUMMARIZE = auto()
 384    TABLE_SAMPLE = auto()
 385    TAG = auto()
 386    TEMPORARY = auto()
 387    TOP = auto()
 388    THEN = auto()
 389    TRUE = auto()
 390    TRUNCATE = auto()
 391    UNCACHE = auto()
 392    UNION = auto()
 393    UNNEST = auto()
 394    UNPIVOT = auto()
 395    UPDATE = auto()
 396    USE = auto()
 397    USING = auto()
 398    VALUES = auto()
 399    VIEW = auto()
 400    VOLATILE = auto()
 401    WHEN = auto()
 402    WHERE = auto()
 403    WINDOW = auto()
 404    WITH = auto()
 405    UNIQUE = auto()
 406    VERSION_SNAPSHOT = auto()
 407    TIMESTAMP_SNAPSHOT = auto()
 408    OPTION = auto()
 409    SINK = auto()
 410    SOURCE = auto()
 411
 412
 413_ALL_TOKEN_TYPES = list(TokenType)
 414_TOKEN_TYPE_TO_INDEX = {token_type: i for i, token_type in enumerate(_ALL_TOKEN_TYPES)}
 415
 416
 417class Token:
 418    __slots__ = ("token_type", "text", "line", "col", "start", "end", "comments")
 419
 420    @classmethod
 421    def number(cls, number: int) -> Token:
 422        """Returns a NUMBER token with `number` as its text."""
 423        return cls(TokenType.NUMBER, str(number))
 424
 425    @classmethod
 426    def string(cls, string: str) -> Token:
 427        """Returns a STRING token with `string` as its text."""
 428        return cls(TokenType.STRING, string)
 429
 430    @classmethod
 431    def identifier(cls, identifier: str) -> Token:
 432        """Returns an IDENTIFIER token with `identifier` as its text."""
 433        return cls(TokenType.IDENTIFIER, identifier)
 434
 435    @classmethod
 436    def var(cls, var: str) -> Token:
 437        """Returns an VAR token with `var` as its text."""
 438        return cls(TokenType.VAR, var)
 439
 440    def __init__(
 441        self,
 442        token_type: TokenType,
 443        text: str,
 444        line: int = 1,
 445        col: int = 1,
 446        start: int = 0,
 447        end: int = 0,
 448        comments: t.Optional[t.List[str]] = None,
 449    ) -> None:
 450        """Token initializer.
 451
 452        Args:
 453            token_type: The TokenType Enum.
 454            text: The text of the token.
 455            line: The line that the token ends on.
 456            col: The column that the token ends on.
 457            start: The start index of the token.
 458            end: The ending index of the token.
 459            comments: The comments to attach to the token.
 460        """
 461        self.token_type = token_type
 462        self.text = text
 463        self.line = line
 464        self.col = col
 465        self.start = start
 466        self.end = end
 467        self.comments = [] if comments is None else comments
 468
 469    def __repr__(self) -> str:
 470        attributes = ", ".join(f"{k}: {getattr(self, k)}" for k in self.__slots__)
 471        return f"<Token {attributes}>"
 472
 473
 474class _Tokenizer(type):
 475    def __new__(cls, clsname, bases, attrs):
 476        klass = super().__new__(cls, clsname, bases, attrs)
 477
 478        def _convert_quotes(arr: t.List[str | t.Tuple[str, str]]) -> t.Dict[str, str]:
 479            return dict(
 480                (item, item) if isinstance(item, str) else (item[0], item[1]) for item in arr
 481            )
 482
 483        def _quotes_to_format(
 484            token_type: TokenType, arr: t.List[str | t.Tuple[str, str]]
 485        ) -> t.Dict[str, t.Tuple[str, TokenType]]:
 486            return {k: (v, token_type) for k, v in _convert_quotes(arr).items()}
 487
 488        klass._QUOTES = _convert_quotes(klass.QUOTES)
 489        klass._IDENTIFIERS = _convert_quotes(klass.IDENTIFIERS)
 490
 491        klass._FORMAT_STRINGS = {
 492            **{
 493                p + s: (e, TokenType.NATIONAL_STRING)
 494                for s, e in klass._QUOTES.items()
 495                for p in ("n", "N")
 496            },
 497            **_quotes_to_format(TokenType.BIT_STRING, klass.BIT_STRINGS),
 498            **_quotes_to_format(TokenType.BYTE_STRING, klass.BYTE_STRINGS),
 499            **_quotes_to_format(TokenType.HEX_STRING, klass.HEX_STRINGS),
 500            **_quotes_to_format(TokenType.RAW_STRING, klass.RAW_STRINGS),
 501            **_quotes_to_format(TokenType.HEREDOC_STRING, klass.HEREDOC_STRINGS),
 502            **_quotes_to_format(TokenType.UNICODE_STRING, klass.UNICODE_STRINGS),
 503        }
 504
 505        klass._STRING_ESCAPES = set(klass.STRING_ESCAPES)
 506        klass._IDENTIFIER_ESCAPES = set(klass.IDENTIFIER_ESCAPES)
 507        klass._COMMENTS = {
 508            **dict(
 509                (comment, None) if isinstance(comment, str) else (comment[0], comment[1])
 510                for comment in klass.COMMENTS
 511            ),
 512            "{#": "#}",  # Ensure Jinja comments are tokenized correctly in all dialects
 513        }
 514        if klass.HINT_START in klass.KEYWORDS:
 515            klass._COMMENTS[klass.HINT_START] = "*/"
 516
 517        klass._KEYWORD_TRIE = new_trie(
 518            key.upper()
 519            for key in (
 520                *klass.KEYWORDS,
 521                *klass._COMMENTS,
 522                *klass._QUOTES,
 523                *klass._FORMAT_STRINGS,
 524            )
 525            if " " in key or any(single in key for single in klass.SINGLE_TOKENS)
 526        )
 527
 528        if USE_RS_TOKENIZER:
 529            settings = RsTokenizerSettings(
 530                white_space={k: _TOKEN_TYPE_TO_INDEX[v] for k, v in klass.WHITE_SPACE.items()},
 531                single_tokens={k: _TOKEN_TYPE_TO_INDEX[v] for k, v in klass.SINGLE_TOKENS.items()},
 532                keywords={k: _TOKEN_TYPE_TO_INDEX[v] for k, v in klass.KEYWORDS.items()},
 533                numeric_literals=klass.NUMERIC_LITERALS,
 534                identifiers=klass._IDENTIFIERS,
 535                identifier_escapes=klass._IDENTIFIER_ESCAPES,
 536                string_escapes=klass._STRING_ESCAPES,
 537                quotes=klass._QUOTES,
 538                format_strings={
 539                    k: (v1, _TOKEN_TYPE_TO_INDEX[v2])
 540                    for k, (v1, v2) in klass._FORMAT_STRINGS.items()
 541                },
 542                has_bit_strings=bool(klass.BIT_STRINGS),
 543                has_hex_strings=bool(klass.HEX_STRINGS),
 544                comments=klass._COMMENTS,
 545                var_single_tokens=klass.VAR_SINGLE_TOKENS,
 546                commands={_TOKEN_TYPE_TO_INDEX[v] for v in klass.COMMANDS},
 547                command_prefix_tokens={
 548                    _TOKEN_TYPE_TO_INDEX[v] for v in klass.COMMAND_PREFIX_TOKENS
 549                },
 550                heredoc_tag_is_identifier=klass.HEREDOC_TAG_IS_IDENTIFIER,
 551                string_escapes_allowed_in_raw_strings=klass.STRING_ESCAPES_ALLOWED_IN_RAW_STRINGS,
 552                nested_comments=klass.NESTED_COMMENTS,
 553                hint_start=klass.HINT_START,
 554                tokens_preceding_hint={
 555                    _TOKEN_TYPE_TO_INDEX[v] for v in klass.TOKENS_PRECEDING_HINT
 556                },
 557            )
 558            token_types = RsTokenTypeSettings(
 559                bit_string=_TOKEN_TYPE_TO_INDEX[TokenType.BIT_STRING],
 560                break_=_TOKEN_TYPE_TO_INDEX[TokenType.BREAK],
 561                dcolon=_TOKEN_TYPE_TO_INDEX[TokenType.DCOLON],
 562                heredoc_string=_TOKEN_TYPE_TO_INDEX[TokenType.HEREDOC_STRING],
 563                raw_string=_TOKEN_TYPE_TO_INDEX[TokenType.RAW_STRING],
 564                hex_string=_TOKEN_TYPE_TO_INDEX[TokenType.HEX_STRING],
 565                identifier=_TOKEN_TYPE_TO_INDEX[TokenType.IDENTIFIER],
 566                number=_TOKEN_TYPE_TO_INDEX[TokenType.NUMBER],
 567                parameter=_TOKEN_TYPE_TO_INDEX[TokenType.PARAMETER],
 568                semicolon=_TOKEN_TYPE_TO_INDEX[TokenType.SEMICOLON],
 569                string=_TOKEN_TYPE_TO_INDEX[TokenType.STRING],
 570                var=_TOKEN_TYPE_TO_INDEX[TokenType.VAR],
 571                heredoc_string_alternative=_TOKEN_TYPE_TO_INDEX[klass.HEREDOC_STRING_ALTERNATIVE],
 572                hint=_TOKEN_TYPE_TO_INDEX[TokenType.HINT],
 573            )
 574            klass._RS_TOKENIZER = RsTokenizer(settings, token_types)
 575        else:
 576            klass._RS_TOKENIZER = None
 577
 578        return klass
 579
 580
 581class Tokenizer(metaclass=_Tokenizer):
 582    SINGLE_TOKENS = {
 583        "(": TokenType.L_PAREN,
 584        ")": TokenType.R_PAREN,
 585        "[": TokenType.L_BRACKET,
 586        "]": TokenType.R_BRACKET,
 587        "{": TokenType.L_BRACE,
 588        "}": TokenType.R_BRACE,
 589        "&": TokenType.AMP,
 590        "^": TokenType.CARET,
 591        ":": TokenType.COLON,
 592        ",": TokenType.COMMA,
 593        ".": TokenType.DOT,
 594        "-": TokenType.DASH,
 595        "=": TokenType.EQ,
 596        ">": TokenType.GT,
 597        "<": TokenType.LT,
 598        "%": TokenType.MOD,
 599        "!": TokenType.NOT,
 600        "|": TokenType.PIPE,
 601        "+": TokenType.PLUS,
 602        ";": TokenType.SEMICOLON,
 603        "/": TokenType.SLASH,
 604        "\\": TokenType.BACKSLASH,
 605        "*": TokenType.STAR,
 606        "~": TokenType.TILDA,
 607        "?": TokenType.PLACEHOLDER,
 608        "@": TokenType.PARAMETER,
 609        "#": TokenType.HASH,
 610        # Used for breaking a var like x'y' but nothing else the token type doesn't matter
 611        "'": TokenType.UNKNOWN,
 612        "`": TokenType.UNKNOWN,
 613        '"': TokenType.UNKNOWN,
 614    }
 615
 616    BIT_STRINGS: t.List[str | t.Tuple[str, str]] = []
 617    BYTE_STRINGS: t.List[str | t.Tuple[str, str]] = []
 618    HEX_STRINGS: t.List[str | t.Tuple[str, str]] = []
 619    RAW_STRINGS: t.List[str | t.Tuple[str, str]] = []
 620    HEREDOC_STRINGS: t.List[str | t.Tuple[str, str]] = []
 621    UNICODE_STRINGS: t.List[str | t.Tuple[str, str]] = []
 622    IDENTIFIERS: t.List[str | t.Tuple[str, str]] = ['"']
 623    QUOTES: t.List[t.Tuple[str, str] | str] = ["'"]
 624    STRING_ESCAPES = ["'"]
 625    VAR_SINGLE_TOKENS: t.Set[str] = set()
 626
 627    # The strings in this list can always be used as escapes, regardless of the surrounding
 628    # identifier delimiters. By default, the closing delimiter is assumed to also act as an
 629    # identifier escape, e.g. if we use double-quotes, then they also act as escapes: "x"""
 630    IDENTIFIER_ESCAPES: t.List[str] = []
 631
 632    # Whether the heredoc tags follow the same lexical rules as unquoted identifiers
 633    HEREDOC_TAG_IS_IDENTIFIER = False
 634
 635    # Token that we'll generate as a fallback if the heredoc prefix doesn't correspond to a heredoc
 636    HEREDOC_STRING_ALTERNATIVE = TokenType.VAR
 637
 638    # Whether string escape characters function as such when placed within raw strings
 639    STRING_ESCAPES_ALLOWED_IN_RAW_STRINGS = True
 640
 641    NESTED_COMMENTS = True
 642
 643    HINT_START = "/*+"
 644
 645    TOKENS_PRECEDING_HINT = {TokenType.SELECT, TokenType.INSERT, TokenType.UPDATE, TokenType.DELETE}
 646
 647    # Autofilled
 648    _COMMENTS: t.Dict[str, str] = {}
 649    _FORMAT_STRINGS: t.Dict[str, t.Tuple[str, TokenType]] = {}
 650    _IDENTIFIERS: t.Dict[str, str] = {}
 651    _IDENTIFIER_ESCAPES: t.Set[str] = set()
 652    _QUOTES: t.Dict[str, str] = {}
 653    _STRING_ESCAPES: t.Set[str] = set()
 654    _KEYWORD_TRIE: t.Dict = {}
 655    _RS_TOKENIZER: t.Optional[t.Any] = None
 656
 657    KEYWORDS: t.Dict[str, TokenType] = {
 658        **{f"{{%{postfix}": TokenType.BLOCK_START for postfix in ("", "+", "-")},
 659        **{f"{prefix}%}}": TokenType.BLOCK_END for prefix in ("", "+", "-")},
 660        **{f"{{{{{postfix}": TokenType.BLOCK_START for postfix in ("+", "-")},
 661        **{f"{prefix}}}}}": TokenType.BLOCK_END for prefix in ("+", "-")},
 662        HINT_START: TokenType.HINT,
 663        "==": TokenType.EQ,
 664        "::": TokenType.DCOLON,
 665        "||": TokenType.DPIPE,
 666        ">=": TokenType.GTE,
 667        "<=": TokenType.LTE,
 668        "<>": TokenType.NEQ,
 669        "!=": TokenType.NEQ,
 670        ":=": TokenType.COLON_EQ,
 671        "<=>": TokenType.NULLSAFE_EQ,
 672        "->": TokenType.ARROW,
 673        "->>": TokenType.DARROW,
 674        "=>": TokenType.FARROW,
 675        "#>": TokenType.HASH_ARROW,
 676        "#>>": TokenType.DHASH_ARROW,
 677        "<->": TokenType.LR_ARROW,
 678        "&&": TokenType.DAMP,
 679        "??": TokenType.DQMARK,
 680        "~~~": TokenType.GLOB,
 681        "~~": TokenType.LIKE,
 682        "~~*": TokenType.ILIKE,
 683        "~*": TokenType.IRLIKE,
 684        "ALL": TokenType.ALL,
 685        "ALWAYS": TokenType.ALWAYS,
 686        "AND": TokenType.AND,
 687        "ANTI": TokenType.ANTI,
 688        "ANY": TokenType.ANY,
 689        "ASC": TokenType.ASC,
 690        "AS": TokenType.ALIAS,
 691        "ASOF": TokenType.ASOF,
 692        "AUTOINCREMENT": TokenType.AUTO_INCREMENT,
 693        "AUTO_INCREMENT": TokenType.AUTO_INCREMENT,
 694        "BEGIN": TokenType.BEGIN,
 695        "BETWEEN": TokenType.BETWEEN,
 696        "CACHE": TokenType.CACHE,
 697        "UNCACHE": TokenType.UNCACHE,
 698        "CASE": TokenType.CASE,
 699        "CHARACTER SET": TokenType.CHARACTER_SET,
 700        "CLUSTER BY": TokenType.CLUSTER_BY,
 701        "COLLATE": TokenType.COLLATE,
 702        "COLUMN": TokenType.COLUMN,
 703        "COMMIT": TokenType.COMMIT,
 704        "CONNECT BY": TokenType.CONNECT_BY,
 705        "CONSTRAINT": TokenType.CONSTRAINT,
 706        "COPY": TokenType.COPY,
 707        "CREATE": TokenType.CREATE,
 708        "CROSS": TokenType.CROSS,
 709        "CUBE": TokenType.CUBE,
 710        "CURRENT_DATE": TokenType.CURRENT_DATE,
 711        "CURRENT_TIME": TokenType.CURRENT_TIME,
 712        "CURRENT_TIMESTAMP": TokenType.CURRENT_TIMESTAMP,
 713        "CURRENT_USER": TokenType.CURRENT_USER,
 714        "DATABASE": TokenType.DATABASE,
 715        "DEFAULT": TokenType.DEFAULT,
 716        "DELETE": TokenType.DELETE,
 717        "DESC": TokenType.DESC,
 718        "DESCRIBE": TokenType.DESCRIBE,
 719        "DISTINCT": TokenType.DISTINCT,
 720        "DISTRIBUTE BY": TokenType.DISTRIBUTE_BY,
 721        "DIV": TokenType.DIV,
 722        "DROP": TokenType.DROP,
 723        "ELSE": TokenType.ELSE,
 724        "END": TokenType.END,
 725        "ENUM": TokenType.ENUM,
 726        "ESCAPE": TokenType.ESCAPE,
 727        "EXCEPT": TokenType.EXCEPT,
 728        "EXECUTE": TokenType.EXECUTE,
 729        "EXISTS": TokenType.EXISTS,
 730        "FALSE": TokenType.FALSE,
 731        "FETCH": TokenType.FETCH,
 732        "FILTER": TokenType.FILTER,
 733        "FIRST": TokenType.FIRST,
 734        "FULL": TokenType.FULL,
 735        "FUNCTION": TokenType.FUNCTION,
 736        "FOR": TokenType.FOR,
 737        "FOREIGN KEY": TokenType.FOREIGN_KEY,
 738        "FORMAT": TokenType.FORMAT,
 739        "FROM": TokenType.FROM,
 740        "GEOGRAPHY": TokenType.GEOGRAPHY,
 741        "GEOMETRY": TokenType.GEOMETRY,
 742        "GLOB": TokenType.GLOB,
 743        "GROUP BY": TokenType.GROUP_BY,
 744        "GROUPING SETS": TokenType.GROUPING_SETS,
 745        "HAVING": TokenType.HAVING,
 746        "ILIKE": TokenType.ILIKE,
 747        "IN": TokenType.IN,
 748        "INDEX": TokenType.INDEX,
 749        "INET": TokenType.INET,
 750        "INNER": TokenType.INNER,
 751        "INSERT": TokenType.INSERT,
 752        "INTERVAL": TokenType.INTERVAL,
 753        "INTERSECT": TokenType.INTERSECT,
 754        "INTO": TokenType.INTO,
 755        "IS": TokenType.IS,
 756        "ISNULL": TokenType.ISNULL,
 757        "JOIN": TokenType.JOIN,
 758        "KEEP": TokenType.KEEP,
 759        "KILL": TokenType.KILL,
 760        "LATERAL": TokenType.LATERAL,
 761        "LEFT": TokenType.LEFT,
 762        "LIKE": TokenType.LIKE,
 763        "LIMIT": TokenType.LIMIT,
 764        "LOAD": TokenType.LOAD,
 765        "LOCK": TokenType.LOCK,
 766        "MERGE": TokenType.MERGE,
 767        "NATURAL": TokenType.NATURAL,
 768        "NEXT": TokenType.NEXT,
 769        "NOT": TokenType.NOT,
 770        "NOTNULL": TokenType.NOTNULL,
 771        "NULL": TokenType.NULL,
 772        "OBJECT": TokenType.OBJECT,
 773        "OFFSET": TokenType.OFFSET,
 774        "ON": TokenType.ON,
 775        "OR": TokenType.OR,
 776        "XOR": TokenType.XOR,
 777        "ORDER BY": TokenType.ORDER_BY,
 778        "ORDINALITY": TokenType.ORDINALITY,
 779        "OUTER": TokenType.OUTER,
 780        "OVER": TokenType.OVER,
 781        "OVERLAPS": TokenType.OVERLAPS,
 782        "OVERWRITE": TokenType.OVERWRITE,
 783        "PARTITION": TokenType.PARTITION,
 784        "PARTITION BY": TokenType.PARTITION_BY,
 785        "PARTITIONED BY": TokenType.PARTITION_BY,
 786        "PARTITIONED_BY": TokenType.PARTITION_BY,
 787        "PERCENT": TokenType.PERCENT,
 788        "PIVOT": TokenType.PIVOT,
 789        "PRAGMA": TokenType.PRAGMA,
 790        "PRIMARY KEY": TokenType.PRIMARY_KEY,
 791        "PROCEDURE": TokenType.PROCEDURE,
 792        "QUALIFY": TokenType.QUALIFY,
 793        "RANGE": TokenType.RANGE,
 794        "RECURSIVE": TokenType.RECURSIVE,
 795        "REGEXP": TokenType.RLIKE,
 796        "RENAME": TokenType.RENAME,
 797        "REPLACE": TokenType.REPLACE,
 798        "RETURNING": TokenType.RETURNING,
 799        "REFERENCES": TokenType.REFERENCES,
 800        "RIGHT": TokenType.RIGHT,
 801        "RLIKE": TokenType.RLIKE,
 802        "ROLLBACK": TokenType.ROLLBACK,
 803        "ROLLUP": TokenType.ROLLUP,
 804        "ROW": TokenType.ROW,
 805        "ROWS": TokenType.ROWS,
 806        "SCHEMA": TokenType.SCHEMA,
 807        "SELECT": TokenType.SELECT,
 808        "SEMI": TokenType.SEMI,
 809        "SET": TokenType.SET,
 810        "SETTINGS": TokenType.SETTINGS,
 811        "SHOW": TokenType.SHOW,
 812        "SIMILAR TO": TokenType.SIMILAR_TO,
 813        "SOME": TokenType.SOME,
 814        "SORT BY": TokenType.SORT_BY,
 815        "START WITH": TokenType.START_WITH,
 816        "STRAIGHT_JOIN": TokenType.STRAIGHT_JOIN,
 817        "TABLE": TokenType.TABLE,
 818        "TABLESAMPLE": TokenType.TABLE_SAMPLE,
 819        "TEMP": TokenType.TEMPORARY,
 820        "TEMPORARY": TokenType.TEMPORARY,
 821        "THEN": TokenType.THEN,
 822        "TRUE": TokenType.TRUE,
 823        "TRUNCATE": TokenType.TRUNCATE,
 824        "UNION": TokenType.UNION,
 825        "UNKNOWN": TokenType.UNKNOWN,
 826        "UNNEST": TokenType.UNNEST,
 827        "UNPIVOT": TokenType.UNPIVOT,
 828        "UPDATE": TokenType.UPDATE,
 829        "USE": TokenType.USE,
 830        "USING": TokenType.USING,
 831        "UUID": TokenType.UUID,
 832        "VALUES": TokenType.VALUES,
 833        "VIEW": TokenType.VIEW,
 834        "VOLATILE": TokenType.VOLATILE,
 835        "WHEN": TokenType.WHEN,
 836        "WHERE": TokenType.WHERE,
 837        "WINDOW": TokenType.WINDOW,
 838        "WITH": TokenType.WITH,
 839        "APPLY": TokenType.APPLY,
 840        "ARRAY": TokenType.ARRAY,
 841        "BIT": TokenType.BIT,
 842        "BOOL": TokenType.BOOLEAN,
 843        "BOOLEAN": TokenType.BOOLEAN,
 844        "BYTE": TokenType.TINYINT,
 845        "MEDIUMINT": TokenType.MEDIUMINT,
 846        "INT1": TokenType.TINYINT,
 847        "TINYINT": TokenType.TINYINT,
 848        "INT16": TokenType.SMALLINT,
 849        "SHORT": TokenType.SMALLINT,
 850        "SMALLINT": TokenType.SMALLINT,
 851        "INT128": TokenType.INT128,
 852        "HUGEINT": TokenType.INT128,
 853        "UHUGEINT": TokenType.UINT128,
 854        "INT2": TokenType.SMALLINT,
 855        "INTEGER": TokenType.INT,
 856        "INT": TokenType.INT,
 857        "INT4": TokenType.INT,
 858        "INT32": TokenType.INT,
 859        "INT64": TokenType.BIGINT,
 860        "LONG": TokenType.BIGINT,
 861        "BIGINT": TokenType.BIGINT,
 862        "INT8": TokenType.TINYINT,
 863        "UINT": TokenType.UINT,
 864        "DEC": TokenType.DECIMAL,
 865        "DECIMAL": TokenType.DECIMAL,
 866        "DECIMAL32": TokenType.DECIMAL32,
 867        "DECIMAL64": TokenType.DECIMAL64,
 868        "DECIMAL128": TokenType.DECIMAL128,
 869        "DECIMAL256": TokenType.DECIMAL256,
 870        "BIGDECIMAL": TokenType.BIGDECIMAL,
 871        "BIGNUMERIC": TokenType.BIGDECIMAL,
 872        "LIST": TokenType.LIST,
 873        "MAP": TokenType.MAP,
 874        "NULLABLE": TokenType.NULLABLE,
 875        "NUMBER": TokenType.DECIMAL,
 876        "NUMERIC": TokenType.DECIMAL,
 877        "FIXED": TokenType.DECIMAL,
 878        "REAL": TokenType.FLOAT,
 879        "FLOAT": TokenType.FLOAT,
 880        "FLOAT4": TokenType.FLOAT,
 881        "FLOAT8": TokenType.DOUBLE,
 882        "DOUBLE": TokenType.DOUBLE,
 883        "DOUBLE PRECISION": TokenType.DOUBLE,
 884        "JSON": TokenType.JSON,
 885        "JSONB": TokenType.JSONB,
 886        "CHAR": TokenType.CHAR,
 887        "CHARACTER": TokenType.CHAR,
 888        "NCHAR": TokenType.NCHAR,
 889        "VARCHAR": TokenType.VARCHAR,
 890        "VARCHAR2": TokenType.VARCHAR,
 891        "NVARCHAR": TokenType.NVARCHAR,
 892        "NVARCHAR2": TokenType.NVARCHAR,
 893        "BPCHAR": TokenType.BPCHAR,
 894        "STR": TokenType.TEXT,
 895        "STRING": TokenType.TEXT,
 896        "TEXT": TokenType.TEXT,
 897        "LONGTEXT": TokenType.LONGTEXT,
 898        "MEDIUMTEXT": TokenType.MEDIUMTEXT,
 899        "TINYTEXT": TokenType.TINYTEXT,
 900        "CLOB": TokenType.TEXT,
 901        "LONGVARCHAR": TokenType.TEXT,
 902        "BINARY": TokenType.BINARY,
 903        "BLOB": TokenType.VARBINARY,
 904        "LONGBLOB": TokenType.LONGBLOB,
 905        "MEDIUMBLOB": TokenType.MEDIUMBLOB,
 906        "TINYBLOB": TokenType.TINYBLOB,
 907        "BYTEA": TokenType.VARBINARY,
 908        "VARBINARY": TokenType.VARBINARY,
 909        "TIME": TokenType.TIME,
 910        "TIMETZ": TokenType.TIMETZ,
 911        "TIMESTAMP": TokenType.TIMESTAMP,
 912        "TIMESTAMPTZ": TokenType.TIMESTAMPTZ,
 913        "TIMESTAMPLTZ": TokenType.TIMESTAMPLTZ,
 914        "TIMESTAMP_LTZ": TokenType.TIMESTAMPLTZ,
 915        "TIMESTAMPNTZ": TokenType.TIMESTAMPNTZ,
 916        "TIMESTAMP_NTZ": TokenType.TIMESTAMPNTZ,
 917        "DATE": TokenType.DATE,
 918        "DATETIME": TokenType.DATETIME,
 919        "INT4RANGE": TokenType.INT4RANGE,
 920        "INT4MULTIRANGE": TokenType.INT4MULTIRANGE,
 921        "INT8RANGE": TokenType.INT8RANGE,
 922        "INT8MULTIRANGE": TokenType.INT8MULTIRANGE,
 923        "NUMRANGE": TokenType.NUMRANGE,
 924        "NUMMULTIRANGE": TokenType.NUMMULTIRANGE,
 925        "TSRANGE": TokenType.TSRANGE,
 926        "TSMULTIRANGE": TokenType.TSMULTIRANGE,
 927        "TSTZRANGE": TokenType.TSTZRANGE,
 928        "TSTZMULTIRANGE": TokenType.TSTZMULTIRANGE,
 929        "DATERANGE": TokenType.DATERANGE,
 930        "DATEMULTIRANGE": TokenType.DATEMULTIRANGE,
 931        "UNIQUE": TokenType.UNIQUE,
 932        "VECTOR": TokenType.VECTOR,
 933        "STRUCT": TokenType.STRUCT,
 934        "SEQUENCE": TokenType.SEQUENCE,
 935        "VARIANT": TokenType.VARIANT,
 936        "ALTER": TokenType.ALTER,
 937        "ANALYZE": TokenType.COMMAND,
 938        "CALL": TokenType.COMMAND,
 939        "COMMENT": TokenType.COMMENT,
 940        "EXPLAIN": TokenType.COMMAND,
 941        "GRANT": TokenType.GRANT,
 942        "OPTIMIZE": TokenType.COMMAND,
 943        "PREPARE": TokenType.COMMAND,
 944        "VACUUM": TokenType.COMMAND,
 945        "USER-DEFINED": TokenType.USERDEFINED,
 946        "FOR VERSION": TokenType.VERSION_SNAPSHOT,
 947        "FOR TIMESTAMP": TokenType.TIMESTAMP_SNAPSHOT,
 948    }
 949
 950    WHITE_SPACE: t.Dict[t.Optional[str], TokenType] = {
 951        " ": TokenType.SPACE,
 952        "\t": TokenType.SPACE,
 953        "\n": TokenType.BREAK,
 954        "\r": TokenType.BREAK,
 955    }
 956
 957    COMMANDS = {
 958        TokenType.COMMAND,
 959        TokenType.EXECUTE,
 960        TokenType.FETCH,
 961        TokenType.SHOW,
 962        TokenType.RENAME,
 963    }
 964
 965    COMMAND_PREFIX_TOKENS = {TokenType.SEMICOLON, TokenType.BEGIN}
 966
 967    # Handle numeric literals like in hive (3L = BIGINT)
 968    NUMERIC_LITERALS: t.Dict[str, str] = {}
 969
 970    COMMENTS = ["--", ("/*", "*/")]
 971
 972    __slots__ = (
 973        "sql",
 974        "size",
 975        "tokens",
 976        "dialect",
 977        "_start",
 978        "_current",
 979        "_line",
 980        "_col",
 981        "_comments",
 982        "_char",
 983        "_end",
 984        "_peek",
 985        "_prev_token_line",
 986        "_rs_dialect_settings",
 987    )
 988
 989    def __init__(self, dialect: DialectType = None) -> None:
 990        from sqlglot.dialects import Dialect
 991
 992        self.dialect = Dialect.get_or_raise(dialect)
 993
 994        if USE_RS_TOKENIZER:
 995            self._rs_dialect_settings = RsTokenizerDialectSettings(
 996                unescaped_sequences=self.dialect.UNESCAPED_SEQUENCES,
 997                identifiers_can_start_with_digit=self.dialect.IDENTIFIERS_CAN_START_WITH_DIGIT,
 998            )
 999
1000        self.reset()
1001
1002    def reset(self) -> None:
1003        self.sql = ""
1004        self.size = 0
1005        self.tokens: t.List[Token] = []
1006        self._start = 0
1007        self._current = 0
1008        self._line = 1
1009        self._col = 0
1010        self._comments: t.List[str] = []
1011
1012        self._char = ""
1013        self._end = False
1014        self._peek = ""
1015        self._prev_token_line = -1
1016
1017    def tokenize(self, sql: str) -> t.List[Token]:
1018        """Returns a list of tokens corresponding to the SQL string `sql`."""
1019        if USE_RS_TOKENIZER:
1020            return self.tokenize_rs(sql)
1021
1022        self.reset()
1023        self.sql = sql
1024        self.size = len(sql)
1025
1026        try:
1027            self._scan()
1028        except Exception as e:
1029            start = max(self._current - 50, 0)
1030            end = min(self._current + 50, self.size - 1)
1031            context = self.sql[start:end]
1032            raise TokenError(f"Error tokenizing '{context}'") from e
1033
1034        return self.tokens
1035
1036    def _scan(self, until: t.Optional[t.Callable] = None) -> None:
1037        while self.size and not self._end:
1038            current = self._current
1039
1040            # Skip spaces here rather than iteratively calling advance() for performance reasons
1041            while current < self.size:
1042                char = self.sql[current]
1043
1044                if char.isspace() and (char == " " or char == "\t"):
1045                    current += 1
1046                else:
1047                    break
1048
1049            offset = current - self._current if current > self._current else 1
1050
1051            self._start = current
1052            self._advance(offset)
1053
1054            if not self._char.isspace():
1055                if self._char.isdigit():
1056                    self._scan_number()
1057                elif self._char in self._IDENTIFIERS:
1058                    self._scan_identifier(self._IDENTIFIERS[self._char])
1059                else:
1060                    self._scan_keywords()
1061
1062            if until and until():
1063                break
1064
1065        if self.tokens and self._comments:
1066            self.tokens[-1].comments.extend(self._comments)
1067
1068    def _chars(self, size: int) -> str:
1069        if size == 1:
1070            return self._char
1071
1072        start = self._current - 1
1073        end = start + size
1074
1075        return self.sql[start:end] if end <= self.size else ""
1076
1077    def _advance(self, i: int = 1, alnum: bool = False) -> None:
1078        if self.WHITE_SPACE.get(self._char) is TokenType.BREAK:
1079            # Ensures we don't count an extra line if we get a \r\n line break sequence
1080            if not (self._char == "\r" and self._peek == "\n"):
1081                self._col = 1
1082                self._line += 1
1083        else:
1084            self._col += i
1085
1086        self._current += i
1087        self._end = self._current >= self.size
1088        self._char = self.sql[self._current - 1]
1089        self._peek = "" if self._end else self.sql[self._current]
1090
1091        if alnum and self._char.isalnum():
1092            # Here we use local variables instead of attributes for better performance
1093            _col = self._col
1094            _current = self._current
1095            _end = self._end
1096            _peek = self._peek
1097
1098            while _peek.isalnum():
1099                _col += 1
1100                _current += 1
1101                _end = _current >= self.size
1102                _peek = "" if _end else self.sql[_current]
1103
1104            self._col = _col
1105            self._current = _current
1106            self._end = _end
1107            self._peek = _peek
1108            self._char = self.sql[_current - 1]
1109
1110    @property
1111    def _text(self) -> str:
1112        return self.sql[self._start : self._current]
1113
1114    def _add(self, token_type: TokenType, text: t.Optional[str] = None) -> None:
1115        self._prev_token_line = self._line
1116
1117        if self._comments and token_type == TokenType.SEMICOLON and self.tokens:
1118            self.tokens[-1].comments.extend(self._comments)
1119            self._comments = []
1120
1121        self.tokens.append(
1122            Token(
1123                token_type,
1124                text=self._text if text is None else text,
1125                line=self._line,
1126                col=self._col,
1127                start=self._start,
1128                end=self._current - 1,
1129                comments=self._comments,
1130            )
1131        )
1132        self._comments = []
1133
1134        # If we have either a semicolon or a begin token before the command's token, we'll parse
1135        # whatever follows the command's token as a string
1136        if (
1137            token_type in self.COMMANDS
1138            and self._peek != ";"
1139            and (len(self.tokens) == 1 or self.tokens[-2].token_type in self.COMMAND_PREFIX_TOKENS)
1140        ):
1141            start = self._current
1142            tokens = len(self.tokens)
1143            self._scan(lambda: self._peek == ";")
1144            self.tokens = self.tokens[:tokens]
1145            text = self.sql[start : self._current].strip()
1146            if text:
1147                self._add(TokenType.STRING, text)
1148
1149    def _scan_keywords(self) -> None:
1150        size = 0
1151        word = None
1152        chars = self._text
1153        char = chars
1154        prev_space = False
1155        skip = False
1156        trie = self._KEYWORD_TRIE
1157        single_token = char in self.SINGLE_TOKENS
1158
1159        while chars:
1160            if skip:
1161                result = TrieResult.PREFIX
1162            else:
1163                result, trie = in_trie(trie, char.upper())
1164
1165            if result == TrieResult.FAILED:
1166                break
1167            if result == TrieResult.EXISTS:
1168                word = chars
1169
1170            end = self._current + size
1171            size += 1
1172
1173            if end < self.size:
1174                char = self.sql[end]
1175                single_token = single_token or char in self.SINGLE_TOKENS
1176                is_space = char.isspace()
1177
1178                if not is_space or not prev_space:
1179                    if is_space:
1180                        char = " "
1181                    chars += char
1182                    prev_space = is_space
1183                    skip = False
1184                else:
1185                    skip = True
1186            else:
1187                char = ""
1188                break
1189
1190        if word:
1191            if self._scan_string(word):
1192                return
1193            if self._scan_comment(word):
1194                return
1195            if prev_space or single_token or not char:
1196                self._advance(size - 1)
1197                word = word.upper()
1198                self._add(self.KEYWORDS[word], text=word)
1199                return
1200
1201        if self._char in self.SINGLE_TOKENS:
1202            self._add(self.SINGLE_TOKENS[self._char], text=self._char)
1203            return
1204
1205        self._scan_var()
1206
1207    def _scan_comment(self, comment_start: str) -> bool:
1208        if comment_start not in self._COMMENTS:
1209            return False
1210
1211        comment_start_line = self._line
1212        comment_start_size = len(comment_start)
1213        comment_end = self._COMMENTS[comment_start]
1214
1215        if comment_end:
1216            # Skip the comment's start delimiter
1217            self._advance(comment_start_size)
1218
1219            comment_count = 1
1220            comment_end_size = len(comment_end)
1221
1222            while not self._end:
1223                if self._chars(comment_end_size) == comment_end:
1224                    comment_count -= 1
1225                    if not comment_count:
1226                        break
1227
1228                self._advance(alnum=True)
1229
1230                # Nested comments are allowed by some dialects, e.g. databricks, duckdb, postgres
1231                if (
1232                    self.NESTED_COMMENTS
1233                    and not self._end
1234                    and self._chars(comment_end_size) == comment_start
1235                ):
1236                    self._advance(comment_start_size)
1237                    comment_count += 1
1238
1239            self._comments.append(self._text[comment_start_size : -comment_end_size + 1])
1240            self._advance(comment_end_size - 1)
1241        else:
1242            while not self._end and self.WHITE_SPACE.get(self._peek) is not TokenType.BREAK:
1243                self._advance(alnum=True)
1244            self._comments.append(self._text[comment_start_size:])
1245
1246        if (
1247            comment_start == self.HINT_START
1248            and self.tokens
1249            and self.tokens[-1].token_type in self.TOKENS_PRECEDING_HINT
1250        ):
1251            self._add(TokenType.HINT)
1252
1253        # Leading comment is attached to the succeeding token, whilst trailing comment to the preceding.
1254        # Multiple consecutive comments are preserved by appending them to the current comments list.
1255        if comment_start_line == self._prev_token_line:
1256            self.tokens[-1].comments.extend(self._comments)
1257            self._comments = []
1258            self._prev_token_line = self._line
1259
1260        return True
1261
1262    def _scan_number(self) -> None:
1263        if self._char == "0":
1264            peek = self._peek.upper()
1265            if peek == "B":
1266                return self._scan_bits() if self.BIT_STRINGS else self._add(TokenType.NUMBER)
1267            elif peek == "X":
1268                return self._scan_hex() if self.HEX_STRINGS else self._add(TokenType.NUMBER)
1269
1270        decimal = False
1271        scientific = 0
1272
1273        while True:
1274            if self._peek.isdigit():
1275                self._advance()
1276            elif self._peek == "." and not decimal:
1277                if self.tokens and self.tokens[-1].token_type == TokenType.PARAMETER:
1278                    return self._add(TokenType.NUMBER)
1279                decimal = True
1280                self._advance()
1281            elif self._peek in ("-", "+") and scientific == 1:
1282                scientific += 1
1283                self._advance()
1284            elif self._peek.upper() == "E" and not scientific:
1285                scientific += 1
1286                self._advance()
1287            elif self._peek.isidentifier():
1288                number_text = self._text
1289                literal = ""
1290
1291                while self._peek.strip() and self._peek not in self.SINGLE_TOKENS:
1292                    literal += self._peek
1293                    self._advance()
1294
1295                token_type = self.KEYWORDS.get(self.NUMERIC_LITERALS.get(literal.upper(), ""))
1296
1297                if token_type:
1298                    self._add(TokenType.NUMBER, number_text)
1299                    self._add(TokenType.DCOLON, "::")
1300                    return self._add(token_type, literal)
1301                elif self.dialect.IDENTIFIERS_CAN_START_WITH_DIGIT:
1302                    return self._add(TokenType.VAR)
1303
1304                self._advance(-len(literal))
1305                return self._add(TokenType.NUMBER, number_text)
1306            else:
1307                return self._add(TokenType.NUMBER)
1308
1309    def _scan_bits(self) -> None:
1310        self._advance()
1311        value = self._extract_value()
1312        try:
1313            # If `value` can't be converted to a binary, fallback to tokenizing it as an identifier
1314            int(value, 2)
1315            self._add(TokenType.BIT_STRING, value[2:])  # Drop the 0b
1316        except ValueError:
1317            self._add(TokenType.IDENTIFIER)
1318
1319    def _scan_hex(self) -> None:
1320        self._advance()
1321        value = self._extract_value()
1322        try:
1323            # If `value` can't be converted to a hex, fallback to tokenizing it as an identifier
1324            int(value, 16)
1325            self._add(TokenType.HEX_STRING, value[2:])  # Drop the 0x
1326        except ValueError:
1327            self._add(TokenType.IDENTIFIER)
1328
1329    def _extract_value(self) -> str:
1330        while True:
1331            char = self._peek.strip()
1332            if char and char not in self.SINGLE_TOKENS:
1333                self._advance(alnum=True)
1334            else:
1335                break
1336
1337        return self._text
1338
1339    def _scan_string(self, start: str) -> bool:
1340        base = None
1341        token_type = TokenType.STRING
1342
1343        if start in self._QUOTES:
1344            end = self._QUOTES[start]
1345        elif start in self._FORMAT_STRINGS:
1346            end, token_type = self._FORMAT_STRINGS[start]
1347
1348            if token_type == TokenType.HEX_STRING:
1349                base = 16
1350            elif token_type == TokenType.BIT_STRING:
1351                base = 2
1352            elif token_type == TokenType.HEREDOC_STRING:
1353                self._advance()
1354
1355                if self._char == end:
1356                    tag = ""
1357                else:
1358                    tag = self._extract_string(
1359                        end,
1360                        raw_string=True,
1361                        raise_unmatched=not self.HEREDOC_TAG_IS_IDENTIFIER,
1362                    )
1363
1364                if tag and self.HEREDOC_TAG_IS_IDENTIFIER and (self._end or not tag.isidentifier()):
1365                    if not self._end:
1366                        self._advance(-1)
1367
1368                    self._advance(-len(tag))
1369                    self._add(self.HEREDOC_STRING_ALTERNATIVE)
1370                    return True
1371
1372                end = f"{start}{tag}{end}"
1373        else:
1374            return False
1375
1376        self._advance(len(start))
1377        text = self._extract_string(end, raw_string=token_type == TokenType.RAW_STRING)
1378
1379        if base:
1380            try:
1381                int(text, base)
1382            except Exception:
1383                raise TokenError(
1384                    f"Numeric string contains invalid characters from {self._line}:{self._start}"
1385                )
1386
1387        self._add(token_type, text)
1388        return True
1389
1390    def _scan_identifier(self, identifier_end: str) -> None:
1391        self._advance()
1392        text = self._extract_string(
1393            identifier_end, escapes=self._IDENTIFIER_ESCAPES | {identifier_end}
1394        )
1395        self._add(TokenType.IDENTIFIER, text)
1396
1397    def _scan_var(self) -> None:
1398        while True:
1399            char = self._peek.strip()
1400            if char and (char in self.VAR_SINGLE_TOKENS or char not in self.SINGLE_TOKENS):
1401                self._advance(alnum=True)
1402            else:
1403                break
1404
1405        self._add(
1406            TokenType.VAR
1407            if self.tokens and self.tokens[-1].token_type == TokenType.PARAMETER
1408            else self.KEYWORDS.get(self._text.upper(), TokenType.VAR)
1409        )
1410
1411    def _extract_string(
1412        self,
1413        delimiter: str,
1414        escapes: t.Optional[t.Set[str]] = None,
1415        raw_string: bool = False,
1416        raise_unmatched: bool = True,
1417    ) -> str:
1418        text = ""
1419        delim_size = len(delimiter)
1420        escapes = self._STRING_ESCAPES if escapes is None else escapes
1421
1422        while True:
1423            if (
1424                not raw_string
1425                and self.dialect.UNESCAPED_SEQUENCES
1426                and self._peek
1427                and self._char in self.STRING_ESCAPES
1428            ):
1429                unescaped_sequence = self.dialect.UNESCAPED_SEQUENCES.get(self._char + self._peek)
1430                if unescaped_sequence:
1431                    self._advance(2)
1432                    text += unescaped_sequence
1433                    continue
1434            if (
1435                (self.STRING_ESCAPES_ALLOWED_IN_RAW_STRINGS or not raw_string)
1436                and self._char in escapes
1437                and (self._peek == delimiter or self._peek in escapes)
1438                and (self._char not in self._QUOTES or self._char == self._peek)
1439            ):
1440                if self._peek == delimiter:
1441                    text += self._peek
1442                else:
1443                    text += self._char + self._peek
1444
1445                if self._current + 1 < self.size:
1446                    self._advance(2)
1447                else:
1448                    raise TokenError(f"Missing {delimiter} from {self._line}:{self._current}")
1449            else:
1450                if self._chars(delim_size) == delimiter:
1451                    if delim_size > 1:
1452                        self._advance(delim_size - 1)
1453                    break
1454
1455                if self._end:
1456                    if not raise_unmatched:
1457                        return text + self._char
1458
1459                    raise TokenError(f"Missing {delimiter} from {self._line}:{self._start}")
1460
1461                current = self._current - 1
1462                self._advance(alnum=True)
1463                text += self.sql[current : self._current - 1]
1464
1465        return text
1466
1467    def tokenize_rs(self, sql: str) -> t.List[Token]:
1468        if not self._RS_TOKENIZER:
1469            raise SqlglotError("Rust tokenizer is not available")
1470
1471        try:
1472            tokens = self._RS_TOKENIZER.tokenize(sql, self._rs_dialect_settings)
1473            for token in tokens:
1474                token.token_type = _ALL_TOKEN_TYPES[token.token_type_index]
1475            return tokens
1476        except Exception as e:
1477            raise TokenError(str(e))