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    DATETIME2 = auto()
 161    DATETIME64 = auto()
 162    SMALLDATETIME = auto()
 163    DATE = auto()
 164    DATE32 = auto()
 165    INT4RANGE = auto()
 166    INT4MULTIRANGE = auto()
 167    INT8RANGE = auto()
 168    INT8MULTIRANGE = auto()
 169    NUMRANGE = auto()
 170    NUMMULTIRANGE = auto()
 171    TSRANGE = auto()
 172    TSMULTIRANGE = auto()
 173    TSTZRANGE = auto()
 174    TSTZMULTIRANGE = auto()
 175    DATERANGE = auto()
 176    DATEMULTIRANGE = auto()
 177    UUID = auto()
 178    GEOGRAPHY = auto()
 179    NULLABLE = auto()
 180    GEOMETRY = auto()
 181    POINT = auto()
 182    RING = auto()
 183    LINESTRING = auto()
 184    MULTILINESTRING = auto()
 185    POLYGON = auto()
 186    MULTIPOLYGON = auto()
 187    HLLSKETCH = auto()
 188    HSTORE = auto()
 189    SUPER = auto()
 190    SERIAL = auto()
 191    SMALLSERIAL = auto()
 192    BIGSERIAL = auto()
 193    XML = auto()
 194    YEAR = auto()
 195    UNIQUEIDENTIFIER = auto()
 196    USERDEFINED = auto()
 197    MONEY = auto()
 198    SMALLMONEY = auto()
 199    ROWVERSION = auto()
 200    IMAGE = auto()
 201    VARIANT = auto()
 202    OBJECT = auto()
 203    INET = auto()
 204    IPADDRESS = auto()
 205    IPPREFIX = auto()
 206    IPV4 = auto()
 207    IPV6 = auto()
 208    ENUM = auto()
 209    ENUM8 = auto()
 210    ENUM16 = auto()
 211    FIXEDSTRING = auto()
 212    LOWCARDINALITY = auto()
 213    NESTED = auto()
 214    AGGREGATEFUNCTION = auto()
 215    SIMPLEAGGREGATEFUNCTION = auto()
 216    TDIGEST = auto()
 217    UNKNOWN = auto()
 218    VECTOR = auto()
 219
 220    # keywords
 221    ALIAS = auto()
 222    ALTER = auto()
 223    ALWAYS = auto()
 224    ALL = auto()
 225    ANTI = auto()
 226    ANY = auto()
 227    APPLY = auto()
 228    ARRAY = auto()
 229    ASC = auto()
 230    ASOF = auto()
 231    ATTACH = auto()
 232    AUTO_INCREMENT = auto()
 233    BEGIN = auto()
 234    BETWEEN = auto()
 235    BULK_COLLECT_INTO = auto()
 236    CACHE = auto()
 237    CASE = auto()
 238    CHARACTER_SET = auto()
 239    CLUSTER_BY = auto()
 240    COLLATE = auto()
 241    COMMAND = auto()
 242    COMMENT = auto()
 243    COMMIT = auto()
 244    CONNECT_BY = auto()
 245    CONSTRAINT = auto()
 246    COPY = auto()
 247    CREATE = auto()
 248    CROSS = auto()
 249    CUBE = auto()
 250    CURRENT_DATE = auto()
 251    CURRENT_DATETIME = auto()
 252    CURRENT_TIME = auto()
 253    CURRENT_TIMESTAMP = auto()
 254    CURRENT_USER = auto()
 255    DECLARE = auto()
 256    DEFAULT = auto()
 257    DELETE = auto()
 258    DESC = auto()
 259    DESCRIBE = auto()
 260    DETACH = auto()
 261    DICTIONARY = auto()
 262    DISTINCT = auto()
 263    DISTRIBUTE_BY = auto()
 264    DIV = auto()
 265    DROP = auto()
 266    ELSE = auto()
 267    END = auto()
 268    ESCAPE = auto()
 269    EXCEPT = auto()
 270    EXECUTE = auto()
 271    EXISTS = auto()
 272    FALSE = auto()
 273    FETCH = auto()
 274    FILTER = auto()
 275    FINAL = auto()
 276    FIRST = auto()
 277    FOR = auto()
 278    FORCE = auto()
 279    FOREIGN_KEY = auto()
 280    FORMAT = auto()
 281    FROM = auto()
 282    FULL = auto()
 283    FUNCTION = auto()
 284    GLOB = auto()
 285    GLOBAL = auto()
 286    GRANT = auto()
 287    GROUP_BY = auto()
 288    GROUPING_SETS = auto()
 289    HAVING = auto()
 290    HINT = auto()
 291    IGNORE = auto()
 292    ILIKE = auto()
 293    ILIKE_ANY = auto()
 294    IN = auto()
 295    INDEX = auto()
 296    INNER = auto()
 297    INSERT = auto()
 298    INTERSECT = auto()
 299    INTERVAL = auto()
 300    INTO = auto()
 301    INTRODUCER = auto()
 302    IRLIKE = auto()
 303    IS = auto()
 304    ISNULL = auto()
 305    JOIN = auto()
 306    JOIN_MARKER = auto()
 307    KEEP = auto()
 308    KEY = auto()
 309    KILL = auto()
 310    LANGUAGE = auto()
 311    LATERAL = auto()
 312    LEFT = auto()
 313    LIKE = auto()
 314    LIKE_ANY = auto()
 315    LIMIT = auto()
 316    LIST = auto()
 317    LOAD = auto()
 318    LOCK = auto()
 319    MAP = auto()
 320    MATCH_CONDITION = auto()
 321    MATCH_RECOGNIZE = auto()
 322    MEMBER_OF = auto()
 323    MERGE = auto()
 324    MOD = auto()
 325    MODEL = auto()
 326    NATURAL = auto()
 327    NEXT = auto()
 328    NOTNULL = auto()
 329    NULL = auto()
 330    OBJECT_IDENTIFIER = auto()
 331    OFFSET = auto()
 332    ON = auto()
 333    ONLY = auto()
 334    OPERATOR = auto()
 335    ORDER_BY = auto()
 336    ORDER_SIBLINGS_BY = auto()
 337    ORDERED = auto()
 338    ORDINALITY = auto()
 339    OUTER = auto()
 340    OVER = auto()
 341    OVERLAPS = auto()
 342    OVERWRITE = auto()
 343    PARTITION = auto()
 344    PARTITION_BY = auto()
 345    PERCENT = auto()
 346    PIVOT = auto()
 347    PLACEHOLDER = auto()
 348    POSITIONAL = auto()
 349    PRAGMA = auto()
 350    PREWHERE = auto()
 351    PRIMARY_KEY = auto()
 352    PROCEDURE = auto()
 353    PROPERTIES = auto()
 354    PSEUDO_TYPE = auto()
 355    QUALIFY = auto()
 356    QUOTE = auto()
 357    RANGE = auto()
 358    RECURSIVE = auto()
 359    REFRESH = auto()
 360    RENAME = auto()
 361    REPLACE = auto()
 362    RETURNING = auto()
 363    REFERENCES = auto()
 364    RIGHT = auto()
 365    RLIKE = auto()
 366    ROLLBACK = auto()
 367    ROLLUP = auto()
 368    ROW = auto()
 369    ROWS = auto()
 370    SELECT = auto()
 371    SEMI = auto()
 372    SEPARATOR = auto()
 373    SEQUENCE = auto()
 374    SERDE_PROPERTIES = auto()
 375    SET = auto()
 376    SETTINGS = auto()
 377    SHOW = auto()
 378    SIMILAR_TO = auto()
 379    SOME = auto()
 380    SORT_BY = auto()
 381    START_WITH = auto()
 382    STORAGE_INTEGRATION = auto()
 383    STRAIGHT_JOIN = auto()
 384    STRUCT = auto()
 385    SUMMARIZE = auto()
 386    TABLE_SAMPLE = auto()
 387    TAG = auto()
 388    TEMPORARY = auto()
 389    TOP = auto()
 390    THEN = auto()
 391    TRUE = auto()
 392    TRUNCATE = auto()
 393    UNCACHE = auto()
 394    UNION = auto()
 395    UNNEST = auto()
 396    UNPIVOT = auto()
 397    UPDATE = auto()
 398    USE = auto()
 399    USING = auto()
 400    VALUES = auto()
 401    VIEW = auto()
 402    VOLATILE = auto()
 403    WHEN = auto()
 404    WHERE = auto()
 405    WINDOW = auto()
 406    WITH = auto()
 407    UNIQUE = auto()
 408    VERSION_SNAPSHOT = auto()
 409    TIMESTAMP_SNAPSHOT = auto()
 410    OPTION = auto()
 411    SINK = auto()
 412    SOURCE = auto()
 413
 414
 415_ALL_TOKEN_TYPES = list(TokenType)
 416_TOKEN_TYPE_TO_INDEX = {token_type: i for i, token_type in enumerate(_ALL_TOKEN_TYPES)}
 417
 418
 419class Token:
 420    __slots__ = ("token_type", "text", "line", "col", "start", "end", "comments")
 421
 422    @classmethod
 423    def number(cls, number: int) -> Token:
 424        """Returns a NUMBER token with `number` as its text."""
 425        return cls(TokenType.NUMBER, str(number))
 426
 427    @classmethod
 428    def string(cls, string: str) -> Token:
 429        """Returns a STRING token with `string` as its text."""
 430        return cls(TokenType.STRING, string)
 431
 432    @classmethod
 433    def identifier(cls, identifier: str) -> Token:
 434        """Returns an IDENTIFIER token with `identifier` as its text."""
 435        return cls(TokenType.IDENTIFIER, identifier)
 436
 437    @classmethod
 438    def var(cls, var: str) -> Token:
 439        """Returns an VAR token with `var` as its text."""
 440        return cls(TokenType.VAR, var)
 441
 442    def __init__(
 443        self,
 444        token_type: TokenType,
 445        text: str,
 446        line: int = 1,
 447        col: int = 1,
 448        start: int = 0,
 449        end: int = 0,
 450        comments: t.Optional[t.List[str]] = None,
 451    ) -> None:
 452        """Token initializer.
 453
 454        Args:
 455            token_type: The TokenType Enum.
 456            text: The text of the token.
 457            line: The line that the token ends on.
 458            col: The column that the token ends on.
 459            start: The start index of the token.
 460            end: The ending index of the token.
 461            comments: The comments to attach to the token.
 462        """
 463        self.token_type = token_type
 464        self.text = text
 465        self.line = line
 466        self.col = col
 467        self.start = start
 468        self.end = end
 469        self.comments = [] if comments is None else comments
 470
 471    def __repr__(self) -> str:
 472        attributes = ", ".join(f"{k}: {getattr(self, k)}" for k in self.__slots__)
 473        return f"<Token {attributes}>"
 474
 475
 476class _Tokenizer(type):
 477    def __new__(cls, clsname, bases, attrs):
 478        klass = super().__new__(cls, clsname, bases, attrs)
 479
 480        def _convert_quotes(arr: t.List[str | t.Tuple[str, str]]) -> t.Dict[str, str]:
 481            return dict(
 482                (item, item) if isinstance(item, str) else (item[0], item[1]) for item in arr
 483            )
 484
 485        def _quotes_to_format(
 486            token_type: TokenType, arr: t.List[str | t.Tuple[str, str]]
 487        ) -> t.Dict[str, t.Tuple[str, TokenType]]:
 488            return {k: (v, token_type) for k, v in _convert_quotes(arr).items()}
 489
 490        klass._QUOTES = _convert_quotes(klass.QUOTES)
 491        klass._IDENTIFIERS = _convert_quotes(klass.IDENTIFIERS)
 492
 493        klass._FORMAT_STRINGS = {
 494            **{
 495                p + s: (e, TokenType.NATIONAL_STRING)
 496                for s, e in klass._QUOTES.items()
 497                for p in ("n", "N")
 498            },
 499            **_quotes_to_format(TokenType.BIT_STRING, klass.BIT_STRINGS),
 500            **_quotes_to_format(TokenType.BYTE_STRING, klass.BYTE_STRINGS),
 501            **_quotes_to_format(TokenType.HEX_STRING, klass.HEX_STRINGS),
 502            **_quotes_to_format(TokenType.RAW_STRING, klass.RAW_STRINGS),
 503            **_quotes_to_format(TokenType.HEREDOC_STRING, klass.HEREDOC_STRINGS),
 504            **_quotes_to_format(TokenType.UNICODE_STRING, klass.UNICODE_STRINGS),
 505        }
 506
 507        klass._STRING_ESCAPES = set(klass.STRING_ESCAPES)
 508        klass._IDENTIFIER_ESCAPES = set(klass.IDENTIFIER_ESCAPES)
 509        klass._COMMENTS = {
 510            **dict(
 511                (comment, None) if isinstance(comment, str) else (comment[0], comment[1])
 512                for comment in klass.COMMENTS
 513            ),
 514            "{#": "#}",  # Ensure Jinja comments are tokenized correctly in all dialects
 515        }
 516        if klass.HINT_START in klass.KEYWORDS:
 517            klass._COMMENTS[klass.HINT_START] = "*/"
 518
 519        klass._KEYWORD_TRIE = new_trie(
 520            key.upper()
 521            for key in (
 522                *klass.KEYWORDS,
 523                *klass._COMMENTS,
 524                *klass._QUOTES,
 525                *klass._FORMAT_STRINGS,
 526            )
 527            if " " in key or any(single in key for single in klass.SINGLE_TOKENS)
 528        )
 529
 530        if USE_RS_TOKENIZER:
 531            settings = RsTokenizerSettings(
 532                white_space={k: _TOKEN_TYPE_TO_INDEX[v] for k, v in klass.WHITE_SPACE.items()},
 533                single_tokens={k: _TOKEN_TYPE_TO_INDEX[v] for k, v in klass.SINGLE_TOKENS.items()},
 534                keywords={k: _TOKEN_TYPE_TO_INDEX[v] for k, v in klass.KEYWORDS.items()},
 535                numeric_literals=klass.NUMERIC_LITERALS,
 536                identifiers=klass._IDENTIFIERS,
 537                identifier_escapes=klass._IDENTIFIER_ESCAPES,
 538                string_escapes=klass._STRING_ESCAPES,
 539                quotes=klass._QUOTES,
 540                format_strings={
 541                    k: (v1, _TOKEN_TYPE_TO_INDEX[v2])
 542                    for k, (v1, v2) in klass._FORMAT_STRINGS.items()
 543                },
 544                has_bit_strings=bool(klass.BIT_STRINGS),
 545                has_hex_strings=bool(klass.HEX_STRINGS),
 546                comments=klass._COMMENTS,
 547                var_single_tokens=klass.VAR_SINGLE_TOKENS,
 548                commands={_TOKEN_TYPE_TO_INDEX[v] for v in klass.COMMANDS},
 549                command_prefix_tokens={
 550                    _TOKEN_TYPE_TO_INDEX[v] for v in klass.COMMAND_PREFIX_TOKENS
 551                },
 552                heredoc_tag_is_identifier=klass.HEREDOC_TAG_IS_IDENTIFIER,
 553                string_escapes_allowed_in_raw_strings=klass.STRING_ESCAPES_ALLOWED_IN_RAW_STRINGS,
 554                nested_comments=klass.NESTED_COMMENTS,
 555                hint_start=klass.HINT_START,
 556                tokens_preceding_hint={
 557                    _TOKEN_TYPE_TO_INDEX[v] for v in klass.TOKENS_PRECEDING_HINT
 558                },
 559            )
 560            token_types = RsTokenTypeSettings(
 561                bit_string=_TOKEN_TYPE_TO_INDEX[TokenType.BIT_STRING],
 562                break_=_TOKEN_TYPE_TO_INDEX[TokenType.BREAK],
 563                dcolon=_TOKEN_TYPE_TO_INDEX[TokenType.DCOLON],
 564                heredoc_string=_TOKEN_TYPE_TO_INDEX[TokenType.HEREDOC_STRING],
 565                raw_string=_TOKEN_TYPE_TO_INDEX[TokenType.RAW_STRING],
 566                hex_string=_TOKEN_TYPE_TO_INDEX[TokenType.HEX_STRING],
 567                identifier=_TOKEN_TYPE_TO_INDEX[TokenType.IDENTIFIER],
 568                number=_TOKEN_TYPE_TO_INDEX[TokenType.NUMBER],
 569                parameter=_TOKEN_TYPE_TO_INDEX[TokenType.PARAMETER],
 570                semicolon=_TOKEN_TYPE_TO_INDEX[TokenType.SEMICOLON],
 571                string=_TOKEN_TYPE_TO_INDEX[TokenType.STRING],
 572                var=_TOKEN_TYPE_TO_INDEX[TokenType.VAR],
 573                heredoc_string_alternative=_TOKEN_TYPE_TO_INDEX[klass.HEREDOC_STRING_ALTERNATIVE],
 574                hint=_TOKEN_TYPE_TO_INDEX[TokenType.HINT],
 575            )
 576            klass._RS_TOKENIZER = RsTokenizer(settings, token_types)
 577        else:
 578            klass._RS_TOKENIZER = None
 579
 580        return klass
 581
 582
 583class Tokenizer(metaclass=_Tokenizer):
 584    SINGLE_TOKENS = {
 585        "(": TokenType.L_PAREN,
 586        ")": TokenType.R_PAREN,
 587        "[": TokenType.L_BRACKET,
 588        "]": TokenType.R_BRACKET,
 589        "{": TokenType.L_BRACE,
 590        "}": TokenType.R_BRACE,
 591        "&": TokenType.AMP,
 592        "^": TokenType.CARET,
 593        ":": TokenType.COLON,
 594        ",": TokenType.COMMA,
 595        ".": TokenType.DOT,
 596        "-": TokenType.DASH,
 597        "=": TokenType.EQ,
 598        ">": TokenType.GT,
 599        "<": TokenType.LT,
 600        "%": TokenType.MOD,
 601        "!": TokenType.NOT,
 602        "|": TokenType.PIPE,
 603        "+": TokenType.PLUS,
 604        ";": TokenType.SEMICOLON,
 605        "/": TokenType.SLASH,
 606        "\\": TokenType.BACKSLASH,
 607        "*": TokenType.STAR,
 608        "~": TokenType.TILDA,
 609        "?": TokenType.PLACEHOLDER,
 610        "@": TokenType.PARAMETER,
 611        "#": TokenType.HASH,
 612        # Used for breaking a var like x'y' but nothing else the token type doesn't matter
 613        "'": TokenType.UNKNOWN,
 614        "`": TokenType.UNKNOWN,
 615        '"': TokenType.UNKNOWN,
 616    }
 617
 618    BIT_STRINGS: t.List[str | t.Tuple[str, str]] = []
 619    BYTE_STRINGS: t.List[str | t.Tuple[str, str]] = []
 620    HEX_STRINGS: t.List[str | t.Tuple[str, str]] = []
 621    RAW_STRINGS: t.List[str | t.Tuple[str, str]] = []
 622    HEREDOC_STRINGS: t.List[str | t.Tuple[str, str]] = []
 623    UNICODE_STRINGS: t.List[str | t.Tuple[str, str]] = []
 624    IDENTIFIERS: t.List[str | t.Tuple[str, str]] = ['"']
 625    QUOTES: t.List[t.Tuple[str, str] | str] = ["'"]
 626    STRING_ESCAPES = ["'"]
 627    VAR_SINGLE_TOKENS: t.Set[str] = set()
 628
 629    # The strings in this list can always be used as escapes, regardless of the surrounding
 630    # identifier delimiters. By default, the closing delimiter is assumed to also act as an
 631    # identifier escape, e.g. if we use double-quotes, then they also act as escapes: "x"""
 632    IDENTIFIER_ESCAPES: t.List[str] = []
 633
 634    # Whether the heredoc tags follow the same lexical rules as unquoted identifiers
 635    HEREDOC_TAG_IS_IDENTIFIER = False
 636
 637    # Token that we'll generate as a fallback if the heredoc prefix doesn't correspond to a heredoc
 638    HEREDOC_STRING_ALTERNATIVE = TokenType.VAR
 639
 640    # Whether string escape characters function as such when placed within raw strings
 641    STRING_ESCAPES_ALLOWED_IN_RAW_STRINGS = True
 642
 643    NESTED_COMMENTS = True
 644
 645    HINT_START = "/*+"
 646
 647    TOKENS_PRECEDING_HINT = {TokenType.SELECT, TokenType.INSERT, TokenType.UPDATE, TokenType.DELETE}
 648
 649    # Autofilled
 650    _COMMENTS: t.Dict[str, str] = {}
 651    _FORMAT_STRINGS: t.Dict[str, t.Tuple[str, TokenType]] = {}
 652    _IDENTIFIERS: t.Dict[str, str] = {}
 653    _IDENTIFIER_ESCAPES: t.Set[str] = set()
 654    _QUOTES: t.Dict[str, str] = {}
 655    _STRING_ESCAPES: t.Set[str] = set()
 656    _KEYWORD_TRIE: t.Dict = {}
 657    _RS_TOKENIZER: t.Optional[t.Any] = None
 658
 659    KEYWORDS: t.Dict[str, TokenType] = {
 660        **{f"{{%{postfix}": TokenType.BLOCK_START for postfix in ("", "+", "-")},
 661        **{f"{prefix}%}}": TokenType.BLOCK_END for prefix in ("", "+", "-")},
 662        **{f"{{{{{postfix}": TokenType.BLOCK_START for postfix in ("+", "-")},
 663        **{f"{prefix}}}}}": TokenType.BLOCK_END for prefix in ("+", "-")},
 664        HINT_START: TokenType.HINT,
 665        "==": TokenType.EQ,
 666        "::": TokenType.DCOLON,
 667        "||": TokenType.DPIPE,
 668        ">=": TokenType.GTE,
 669        "<=": TokenType.LTE,
 670        "<>": TokenType.NEQ,
 671        "!=": TokenType.NEQ,
 672        ":=": TokenType.COLON_EQ,
 673        "<=>": TokenType.NULLSAFE_EQ,
 674        "->": TokenType.ARROW,
 675        "->>": TokenType.DARROW,
 676        "=>": TokenType.FARROW,
 677        "#>": TokenType.HASH_ARROW,
 678        "#>>": TokenType.DHASH_ARROW,
 679        "<->": TokenType.LR_ARROW,
 680        "&&": TokenType.DAMP,
 681        "??": TokenType.DQMARK,
 682        "~~~": TokenType.GLOB,
 683        "~~": TokenType.LIKE,
 684        "~~*": TokenType.ILIKE,
 685        "~*": TokenType.IRLIKE,
 686        "ALL": TokenType.ALL,
 687        "ALWAYS": TokenType.ALWAYS,
 688        "AND": TokenType.AND,
 689        "ANTI": TokenType.ANTI,
 690        "ANY": TokenType.ANY,
 691        "ASC": TokenType.ASC,
 692        "AS": TokenType.ALIAS,
 693        "ASOF": TokenType.ASOF,
 694        "AUTOINCREMENT": TokenType.AUTO_INCREMENT,
 695        "AUTO_INCREMENT": TokenType.AUTO_INCREMENT,
 696        "BEGIN": TokenType.BEGIN,
 697        "BETWEEN": TokenType.BETWEEN,
 698        "CACHE": TokenType.CACHE,
 699        "UNCACHE": TokenType.UNCACHE,
 700        "CASE": TokenType.CASE,
 701        "CHARACTER SET": TokenType.CHARACTER_SET,
 702        "CLUSTER BY": TokenType.CLUSTER_BY,
 703        "COLLATE": TokenType.COLLATE,
 704        "COLUMN": TokenType.COLUMN,
 705        "COMMIT": TokenType.COMMIT,
 706        "CONNECT BY": TokenType.CONNECT_BY,
 707        "CONSTRAINT": TokenType.CONSTRAINT,
 708        "COPY": TokenType.COPY,
 709        "CREATE": TokenType.CREATE,
 710        "CROSS": TokenType.CROSS,
 711        "CUBE": TokenType.CUBE,
 712        "CURRENT_DATE": TokenType.CURRENT_DATE,
 713        "CURRENT_TIME": TokenType.CURRENT_TIME,
 714        "CURRENT_TIMESTAMP": TokenType.CURRENT_TIMESTAMP,
 715        "CURRENT_USER": TokenType.CURRENT_USER,
 716        "DATABASE": TokenType.DATABASE,
 717        "DEFAULT": TokenType.DEFAULT,
 718        "DELETE": TokenType.DELETE,
 719        "DESC": TokenType.DESC,
 720        "DESCRIBE": TokenType.DESCRIBE,
 721        "DISTINCT": TokenType.DISTINCT,
 722        "DISTRIBUTE BY": TokenType.DISTRIBUTE_BY,
 723        "DIV": TokenType.DIV,
 724        "DROP": TokenType.DROP,
 725        "ELSE": TokenType.ELSE,
 726        "END": TokenType.END,
 727        "ENUM": TokenType.ENUM,
 728        "ESCAPE": TokenType.ESCAPE,
 729        "EXCEPT": TokenType.EXCEPT,
 730        "EXECUTE": TokenType.EXECUTE,
 731        "EXISTS": TokenType.EXISTS,
 732        "FALSE": TokenType.FALSE,
 733        "FETCH": TokenType.FETCH,
 734        "FILTER": TokenType.FILTER,
 735        "FIRST": TokenType.FIRST,
 736        "FULL": TokenType.FULL,
 737        "FUNCTION": TokenType.FUNCTION,
 738        "FOR": TokenType.FOR,
 739        "FOREIGN KEY": TokenType.FOREIGN_KEY,
 740        "FORMAT": TokenType.FORMAT,
 741        "FROM": TokenType.FROM,
 742        "GEOGRAPHY": TokenType.GEOGRAPHY,
 743        "GEOMETRY": TokenType.GEOMETRY,
 744        "GLOB": TokenType.GLOB,
 745        "GROUP BY": TokenType.GROUP_BY,
 746        "GROUPING SETS": TokenType.GROUPING_SETS,
 747        "HAVING": TokenType.HAVING,
 748        "ILIKE": TokenType.ILIKE,
 749        "IN": TokenType.IN,
 750        "INDEX": TokenType.INDEX,
 751        "INET": TokenType.INET,
 752        "INNER": TokenType.INNER,
 753        "INSERT": TokenType.INSERT,
 754        "INTERVAL": TokenType.INTERVAL,
 755        "INTERSECT": TokenType.INTERSECT,
 756        "INTO": TokenType.INTO,
 757        "IS": TokenType.IS,
 758        "ISNULL": TokenType.ISNULL,
 759        "JOIN": TokenType.JOIN,
 760        "KEEP": TokenType.KEEP,
 761        "KILL": TokenType.KILL,
 762        "LATERAL": TokenType.LATERAL,
 763        "LEFT": TokenType.LEFT,
 764        "LIKE": TokenType.LIKE,
 765        "LIMIT": TokenType.LIMIT,
 766        "LOAD": TokenType.LOAD,
 767        "LOCK": TokenType.LOCK,
 768        "MERGE": TokenType.MERGE,
 769        "NATURAL": TokenType.NATURAL,
 770        "NEXT": TokenType.NEXT,
 771        "NOT": TokenType.NOT,
 772        "NOTNULL": TokenType.NOTNULL,
 773        "NULL": TokenType.NULL,
 774        "OBJECT": TokenType.OBJECT,
 775        "OFFSET": TokenType.OFFSET,
 776        "ON": TokenType.ON,
 777        "OR": TokenType.OR,
 778        "XOR": TokenType.XOR,
 779        "ORDER BY": TokenType.ORDER_BY,
 780        "ORDINALITY": TokenType.ORDINALITY,
 781        "OUTER": TokenType.OUTER,
 782        "OVER": TokenType.OVER,
 783        "OVERLAPS": TokenType.OVERLAPS,
 784        "OVERWRITE": TokenType.OVERWRITE,
 785        "PARTITION": TokenType.PARTITION,
 786        "PARTITION BY": TokenType.PARTITION_BY,
 787        "PARTITIONED BY": TokenType.PARTITION_BY,
 788        "PARTITIONED_BY": TokenType.PARTITION_BY,
 789        "PERCENT": TokenType.PERCENT,
 790        "PIVOT": TokenType.PIVOT,
 791        "PRAGMA": TokenType.PRAGMA,
 792        "PRIMARY KEY": TokenType.PRIMARY_KEY,
 793        "PROCEDURE": TokenType.PROCEDURE,
 794        "QUALIFY": TokenType.QUALIFY,
 795        "RANGE": TokenType.RANGE,
 796        "RECURSIVE": TokenType.RECURSIVE,
 797        "REGEXP": TokenType.RLIKE,
 798        "RENAME": TokenType.RENAME,
 799        "REPLACE": TokenType.REPLACE,
 800        "RETURNING": TokenType.RETURNING,
 801        "REFERENCES": TokenType.REFERENCES,
 802        "RIGHT": TokenType.RIGHT,
 803        "RLIKE": TokenType.RLIKE,
 804        "ROLLBACK": TokenType.ROLLBACK,
 805        "ROLLUP": TokenType.ROLLUP,
 806        "ROW": TokenType.ROW,
 807        "ROWS": TokenType.ROWS,
 808        "SCHEMA": TokenType.SCHEMA,
 809        "SELECT": TokenType.SELECT,
 810        "SEMI": TokenType.SEMI,
 811        "SET": TokenType.SET,
 812        "SETTINGS": TokenType.SETTINGS,
 813        "SHOW": TokenType.SHOW,
 814        "SIMILAR TO": TokenType.SIMILAR_TO,
 815        "SOME": TokenType.SOME,
 816        "SORT BY": TokenType.SORT_BY,
 817        "START WITH": TokenType.START_WITH,
 818        "STRAIGHT_JOIN": TokenType.STRAIGHT_JOIN,
 819        "TABLE": TokenType.TABLE,
 820        "TABLESAMPLE": TokenType.TABLE_SAMPLE,
 821        "TEMP": TokenType.TEMPORARY,
 822        "TEMPORARY": TokenType.TEMPORARY,
 823        "THEN": TokenType.THEN,
 824        "TRUE": TokenType.TRUE,
 825        "TRUNCATE": TokenType.TRUNCATE,
 826        "UNION": TokenType.UNION,
 827        "UNKNOWN": TokenType.UNKNOWN,
 828        "UNNEST": TokenType.UNNEST,
 829        "UNPIVOT": TokenType.UNPIVOT,
 830        "UPDATE": TokenType.UPDATE,
 831        "USE": TokenType.USE,
 832        "USING": TokenType.USING,
 833        "UUID": TokenType.UUID,
 834        "VALUES": TokenType.VALUES,
 835        "VIEW": TokenType.VIEW,
 836        "VOLATILE": TokenType.VOLATILE,
 837        "WHEN": TokenType.WHEN,
 838        "WHERE": TokenType.WHERE,
 839        "WINDOW": TokenType.WINDOW,
 840        "WITH": TokenType.WITH,
 841        "APPLY": TokenType.APPLY,
 842        "ARRAY": TokenType.ARRAY,
 843        "BIT": TokenType.BIT,
 844        "BOOL": TokenType.BOOLEAN,
 845        "BOOLEAN": TokenType.BOOLEAN,
 846        "BYTE": TokenType.TINYINT,
 847        "MEDIUMINT": TokenType.MEDIUMINT,
 848        "INT1": TokenType.TINYINT,
 849        "TINYINT": TokenType.TINYINT,
 850        "INT16": TokenType.SMALLINT,
 851        "SHORT": TokenType.SMALLINT,
 852        "SMALLINT": TokenType.SMALLINT,
 853        "INT128": TokenType.INT128,
 854        "HUGEINT": TokenType.INT128,
 855        "UHUGEINT": TokenType.UINT128,
 856        "INT2": TokenType.SMALLINT,
 857        "INTEGER": TokenType.INT,
 858        "INT": TokenType.INT,
 859        "INT4": TokenType.INT,
 860        "INT32": TokenType.INT,
 861        "INT64": TokenType.BIGINT,
 862        "LONG": TokenType.BIGINT,
 863        "BIGINT": TokenType.BIGINT,
 864        "INT8": TokenType.TINYINT,
 865        "UINT": TokenType.UINT,
 866        "DEC": TokenType.DECIMAL,
 867        "DECIMAL": TokenType.DECIMAL,
 868        "DECIMAL32": TokenType.DECIMAL32,
 869        "DECIMAL64": TokenType.DECIMAL64,
 870        "DECIMAL128": TokenType.DECIMAL128,
 871        "DECIMAL256": TokenType.DECIMAL256,
 872        "BIGDECIMAL": TokenType.BIGDECIMAL,
 873        "BIGNUMERIC": TokenType.BIGDECIMAL,
 874        "LIST": TokenType.LIST,
 875        "MAP": TokenType.MAP,
 876        "NULLABLE": TokenType.NULLABLE,
 877        "NUMBER": TokenType.DECIMAL,
 878        "NUMERIC": TokenType.DECIMAL,
 879        "FIXED": TokenType.DECIMAL,
 880        "REAL": TokenType.FLOAT,
 881        "FLOAT": TokenType.FLOAT,
 882        "FLOAT4": TokenType.FLOAT,
 883        "FLOAT8": TokenType.DOUBLE,
 884        "DOUBLE": TokenType.DOUBLE,
 885        "DOUBLE PRECISION": TokenType.DOUBLE,
 886        "JSON": TokenType.JSON,
 887        "JSONB": TokenType.JSONB,
 888        "CHAR": TokenType.CHAR,
 889        "CHARACTER": TokenType.CHAR,
 890        "NCHAR": TokenType.NCHAR,
 891        "VARCHAR": TokenType.VARCHAR,
 892        "VARCHAR2": TokenType.VARCHAR,
 893        "NVARCHAR": TokenType.NVARCHAR,
 894        "NVARCHAR2": TokenType.NVARCHAR,
 895        "BPCHAR": TokenType.BPCHAR,
 896        "STR": TokenType.TEXT,
 897        "STRING": TokenType.TEXT,
 898        "TEXT": TokenType.TEXT,
 899        "LONGTEXT": TokenType.LONGTEXT,
 900        "MEDIUMTEXT": TokenType.MEDIUMTEXT,
 901        "TINYTEXT": TokenType.TINYTEXT,
 902        "CLOB": TokenType.TEXT,
 903        "LONGVARCHAR": TokenType.TEXT,
 904        "BINARY": TokenType.BINARY,
 905        "BLOB": TokenType.VARBINARY,
 906        "LONGBLOB": TokenType.LONGBLOB,
 907        "MEDIUMBLOB": TokenType.MEDIUMBLOB,
 908        "TINYBLOB": TokenType.TINYBLOB,
 909        "BYTEA": TokenType.VARBINARY,
 910        "VARBINARY": TokenType.VARBINARY,
 911        "TIME": TokenType.TIME,
 912        "TIMETZ": TokenType.TIMETZ,
 913        "TIMESTAMP": TokenType.TIMESTAMP,
 914        "TIMESTAMPTZ": TokenType.TIMESTAMPTZ,
 915        "TIMESTAMPLTZ": TokenType.TIMESTAMPLTZ,
 916        "TIMESTAMP_LTZ": TokenType.TIMESTAMPLTZ,
 917        "TIMESTAMPNTZ": TokenType.TIMESTAMPNTZ,
 918        "TIMESTAMP_NTZ": TokenType.TIMESTAMPNTZ,
 919        "DATE": TokenType.DATE,
 920        "DATETIME": TokenType.DATETIME,
 921        "INT4RANGE": TokenType.INT4RANGE,
 922        "INT4MULTIRANGE": TokenType.INT4MULTIRANGE,
 923        "INT8RANGE": TokenType.INT8RANGE,
 924        "INT8MULTIRANGE": TokenType.INT8MULTIRANGE,
 925        "NUMRANGE": TokenType.NUMRANGE,
 926        "NUMMULTIRANGE": TokenType.NUMMULTIRANGE,
 927        "TSRANGE": TokenType.TSRANGE,
 928        "TSMULTIRANGE": TokenType.TSMULTIRANGE,
 929        "TSTZRANGE": TokenType.TSTZRANGE,
 930        "TSTZMULTIRANGE": TokenType.TSTZMULTIRANGE,
 931        "DATERANGE": TokenType.DATERANGE,
 932        "DATEMULTIRANGE": TokenType.DATEMULTIRANGE,
 933        "UNIQUE": TokenType.UNIQUE,
 934        "VECTOR": TokenType.VECTOR,
 935        "STRUCT": TokenType.STRUCT,
 936        "SEQUENCE": TokenType.SEQUENCE,
 937        "VARIANT": TokenType.VARIANT,
 938        "ALTER": TokenType.ALTER,
 939        "ANALYZE": TokenType.COMMAND,
 940        "CALL": TokenType.COMMAND,
 941        "COMMENT": TokenType.COMMENT,
 942        "EXPLAIN": TokenType.COMMAND,
 943        "GRANT": TokenType.GRANT,
 944        "OPTIMIZE": TokenType.COMMAND,
 945        "PREPARE": TokenType.COMMAND,
 946        "VACUUM": TokenType.COMMAND,
 947        "USER-DEFINED": TokenType.USERDEFINED,
 948        "FOR VERSION": TokenType.VERSION_SNAPSHOT,
 949        "FOR TIMESTAMP": TokenType.TIMESTAMP_SNAPSHOT,
 950    }
 951
 952    WHITE_SPACE: t.Dict[t.Optional[str], TokenType] = {
 953        " ": TokenType.SPACE,
 954        "\t": TokenType.SPACE,
 955        "\n": TokenType.BREAK,
 956        "\r": TokenType.BREAK,
 957    }
 958
 959    COMMANDS = {
 960        TokenType.COMMAND,
 961        TokenType.EXECUTE,
 962        TokenType.FETCH,
 963        TokenType.SHOW,
 964        TokenType.RENAME,
 965    }
 966
 967    COMMAND_PREFIX_TOKENS = {TokenType.SEMICOLON, TokenType.BEGIN}
 968
 969    # Handle numeric literals like in hive (3L = BIGINT)
 970    NUMERIC_LITERALS: t.Dict[str, str] = {}
 971
 972    COMMENTS = ["--", ("/*", "*/")]
 973
 974    __slots__ = (
 975        "sql",
 976        "size",
 977        "tokens",
 978        "dialect",
 979        "_start",
 980        "_current",
 981        "_line",
 982        "_col",
 983        "_comments",
 984        "_char",
 985        "_end",
 986        "_peek",
 987        "_prev_token_line",
 988        "_rs_dialect_settings",
 989    )
 990
 991    def __init__(self, dialect: DialectType = None) -> None:
 992        from sqlglot.dialects import Dialect
 993
 994        self.dialect = Dialect.get_or_raise(dialect)
 995
 996        if USE_RS_TOKENIZER:
 997            self._rs_dialect_settings = RsTokenizerDialectSettings(
 998                unescaped_sequences=self.dialect.UNESCAPED_SEQUENCES,
 999                identifiers_can_start_with_digit=self.dialect.IDENTIFIERS_CAN_START_WITH_DIGIT,
1000            )
1001
1002        self.reset()
1003
1004    def reset(self) -> None:
1005        self.sql = ""
1006        self.size = 0
1007        self.tokens: t.List[Token] = []
1008        self._start = 0
1009        self._current = 0
1010        self._line = 1
1011        self._col = 0
1012        self._comments: t.List[str] = []
1013
1014        self._char = ""
1015        self._end = False
1016        self._peek = ""
1017        self._prev_token_line = -1
1018
1019    def tokenize(self, sql: str) -> t.List[Token]:
1020        """Returns a list of tokens corresponding to the SQL string `sql`."""
1021        if USE_RS_TOKENIZER:
1022            return self.tokenize_rs(sql)
1023
1024        self.reset()
1025        self.sql = sql
1026        self.size = len(sql)
1027
1028        try:
1029            self._scan()
1030        except Exception as e:
1031            start = max(self._current - 50, 0)
1032            end = min(self._current + 50, self.size - 1)
1033            context = self.sql[start:end]
1034            raise TokenError(f"Error tokenizing '{context}'") from e
1035
1036        return self.tokens
1037
1038    def _scan(self, until: t.Optional[t.Callable] = None) -> None:
1039        while self.size and not self._end:
1040            current = self._current
1041
1042            # Skip spaces here rather than iteratively calling advance() for performance reasons
1043            while current < self.size:
1044                char = self.sql[current]
1045
1046                if char.isspace() and (char == " " or char == "\t"):
1047                    current += 1
1048                else:
1049                    break
1050
1051            offset = current - self._current if current > self._current else 1
1052
1053            self._start = current
1054            self._advance(offset)
1055
1056            if not self._char.isspace():
1057                if self._char.isdigit():
1058                    self._scan_number()
1059                elif self._char in self._IDENTIFIERS:
1060                    self._scan_identifier(self._IDENTIFIERS[self._char])
1061                else:
1062                    self._scan_keywords()
1063
1064            if until and until():
1065                break
1066
1067        if self.tokens and self._comments:
1068            self.tokens[-1].comments.extend(self._comments)
1069
1070    def _chars(self, size: int) -> str:
1071        if size == 1:
1072            return self._char
1073
1074        start = self._current - 1
1075        end = start + size
1076
1077        return self.sql[start:end] if end <= self.size else ""
1078
1079    def _advance(self, i: int = 1, alnum: bool = False) -> None:
1080        if self.WHITE_SPACE.get(self._char) is TokenType.BREAK:
1081            # Ensures we don't count an extra line if we get a \r\n line break sequence
1082            if not (self._char == "\r" and self._peek == "\n"):
1083                self._col = 1
1084                self._line += 1
1085        else:
1086            self._col += i
1087
1088        self._current += i
1089        self._end = self._current >= self.size
1090        self._char = self.sql[self._current - 1]
1091        self._peek = "" if self._end else self.sql[self._current]
1092
1093        if alnum and self._char.isalnum():
1094            # Here we use local variables instead of attributes for better performance
1095            _col = self._col
1096            _current = self._current
1097            _end = self._end
1098            _peek = self._peek
1099
1100            while _peek.isalnum():
1101                _col += 1
1102                _current += 1
1103                _end = _current >= self.size
1104                _peek = "" if _end else self.sql[_current]
1105
1106            self._col = _col
1107            self._current = _current
1108            self._end = _end
1109            self._peek = _peek
1110            self._char = self.sql[_current - 1]
1111
1112    @property
1113    def _text(self) -> str:
1114        return self.sql[self._start : self._current]
1115
1116    def _add(self, token_type: TokenType, text: t.Optional[str] = None) -> None:
1117        self._prev_token_line = self._line
1118
1119        if self._comments and token_type == TokenType.SEMICOLON and self.tokens:
1120            self.tokens[-1].comments.extend(self._comments)
1121            self._comments = []
1122
1123        self.tokens.append(
1124            Token(
1125                token_type,
1126                text=self._text if text is None else text,
1127                line=self._line,
1128                col=self._col,
1129                start=self._start,
1130                end=self._current - 1,
1131                comments=self._comments,
1132            )
1133        )
1134        self._comments = []
1135
1136        # If we have either a semicolon or a begin token before the command's token, we'll parse
1137        # whatever follows the command's token as a string
1138        if (
1139            token_type in self.COMMANDS
1140            and self._peek != ";"
1141            and (len(self.tokens) == 1 or self.tokens[-2].token_type in self.COMMAND_PREFIX_TOKENS)
1142        ):
1143            start = self._current
1144            tokens = len(self.tokens)
1145            self._scan(lambda: self._peek == ";")
1146            self.tokens = self.tokens[:tokens]
1147            text = self.sql[start : self._current].strip()
1148            if text:
1149                self._add(TokenType.STRING, text)
1150
1151    def _scan_keywords(self) -> None:
1152        size = 0
1153        word = None
1154        chars = self._text
1155        char = chars
1156        prev_space = False
1157        skip = False
1158        trie = self._KEYWORD_TRIE
1159        single_token = char in self.SINGLE_TOKENS
1160
1161        while chars:
1162            if skip:
1163                result = TrieResult.PREFIX
1164            else:
1165                result, trie = in_trie(trie, char.upper())
1166
1167            if result == TrieResult.FAILED:
1168                break
1169            if result == TrieResult.EXISTS:
1170                word = chars
1171
1172            end = self._current + size
1173            size += 1
1174
1175            if end < self.size:
1176                char = self.sql[end]
1177                single_token = single_token or char in self.SINGLE_TOKENS
1178                is_space = char.isspace()
1179
1180                if not is_space or not prev_space:
1181                    if is_space:
1182                        char = " "
1183                    chars += char
1184                    prev_space = is_space
1185                    skip = False
1186                else:
1187                    skip = True
1188            else:
1189                char = ""
1190                break
1191
1192        if word:
1193            if self._scan_string(word):
1194                return
1195            if self._scan_comment(word):
1196                return
1197            if prev_space or single_token or not char:
1198                self._advance(size - 1)
1199                word = word.upper()
1200                self._add(self.KEYWORDS[word], text=word)
1201                return
1202
1203        if self._char in self.SINGLE_TOKENS:
1204            self._add(self.SINGLE_TOKENS[self._char], text=self._char)
1205            return
1206
1207        self._scan_var()
1208
1209    def _scan_comment(self, comment_start: str) -> bool:
1210        if comment_start not in self._COMMENTS:
1211            return False
1212
1213        comment_start_line = self._line
1214        comment_start_size = len(comment_start)
1215        comment_end = self._COMMENTS[comment_start]
1216
1217        if comment_end:
1218            # Skip the comment's start delimiter
1219            self._advance(comment_start_size)
1220
1221            comment_count = 1
1222            comment_end_size = len(comment_end)
1223
1224            while not self._end:
1225                if self._chars(comment_end_size) == comment_end:
1226                    comment_count -= 1
1227                    if not comment_count:
1228                        break
1229
1230                self._advance(alnum=True)
1231
1232                # Nested comments are allowed by some dialects, e.g. databricks, duckdb, postgres
1233                if (
1234                    self.NESTED_COMMENTS
1235                    and not self._end
1236                    and self._chars(comment_end_size) == comment_start
1237                ):
1238                    self._advance(comment_start_size)
1239                    comment_count += 1
1240
1241            self._comments.append(self._text[comment_start_size : -comment_end_size + 1])
1242            self._advance(comment_end_size - 1)
1243        else:
1244            while not self._end and self.WHITE_SPACE.get(self._peek) is not TokenType.BREAK:
1245                self._advance(alnum=True)
1246            self._comments.append(self._text[comment_start_size:])
1247
1248        if (
1249            comment_start == self.HINT_START
1250            and self.tokens
1251            and self.tokens[-1].token_type in self.TOKENS_PRECEDING_HINT
1252        ):
1253            self._add(TokenType.HINT)
1254
1255        # Leading comment is attached to the succeeding token, whilst trailing comment to the preceding.
1256        # Multiple consecutive comments are preserved by appending them to the current comments list.
1257        if comment_start_line == self._prev_token_line:
1258            self.tokens[-1].comments.extend(self._comments)
1259            self._comments = []
1260            self._prev_token_line = self._line
1261
1262        return True
1263
1264    def _scan_number(self) -> None:
1265        if self._char == "0":
1266            peek = self._peek.upper()
1267            if peek == "B":
1268                return self._scan_bits() if self.BIT_STRINGS else self._add(TokenType.NUMBER)
1269            elif peek == "X":
1270                return self._scan_hex() if self.HEX_STRINGS else self._add(TokenType.NUMBER)
1271
1272        decimal = False
1273        scientific = 0
1274
1275        while True:
1276            if self._peek.isdigit():
1277                self._advance()
1278            elif self._peek == "." and not decimal:
1279                if self.tokens and self.tokens[-1].token_type == TokenType.PARAMETER:
1280                    return self._add(TokenType.NUMBER)
1281                decimal = True
1282                self._advance()
1283            elif self._peek in ("-", "+") and scientific == 1:
1284                scientific += 1
1285                self._advance()
1286            elif self._peek.upper() == "E" and not scientific:
1287                scientific += 1
1288                self._advance()
1289            elif self._peek.isidentifier():
1290                number_text = self._text
1291                literal = ""
1292
1293                while self._peek.strip() and self._peek not in self.SINGLE_TOKENS:
1294                    literal += self._peek
1295                    self._advance()
1296
1297                token_type = self.KEYWORDS.get(self.NUMERIC_LITERALS.get(literal.upper(), ""))
1298
1299                if token_type:
1300                    self._add(TokenType.NUMBER, number_text)
1301                    self._add(TokenType.DCOLON, "::")
1302                    return self._add(token_type, literal)
1303                elif self.dialect.IDENTIFIERS_CAN_START_WITH_DIGIT:
1304                    return self._add(TokenType.VAR)
1305
1306                self._advance(-len(literal))
1307                return self._add(TokenType.NUMBER, number_text)
1308            else:
1309                return self._add(TokenType.NUMBER)
1310
1311    def _scan_bits(self) -> None:
1312        self._advance()
1313        value = self._extract_value()
1314        try:
1315            # If `value` can't be converted to a binary, fallback to tokenizing it as an identifier
1316            int(value, 2)
1317            self._add(TokenType.BIT_STRING, value[2:])  # Drop the 0b
1318        except ValueError:
1319            self._add(TokenType.IDENTIFIER)
1320
1321    def _scan_hex(self) -> None:
1322        self._advance()
1323        value = self._extract_value()
1324        try:
1325            # If `value` can't be converted to a hex, fallback to tokenizing it as an identifier
1326            int(value, 16)
1327            self._add(TokenType.HEX_STRING, value[2:])  # Drop the 0x
1328        except ValueError:
1329            self._add(TokenType.IDENTIFIER)
1330
1331    def _extract_value(self) -> str:
1332        while True:
1333            char = self._peek.strip()
1334            if char and char not in self.SINGLE_TOKENS:
1335                self._advance(alnum=True)
1336            else:
1337                break
1338
1339        return self._text
1340
1341    def _scan_string(self, start: str) -> bool:
1342        base = None
1343        token_type = TokenType.STRING
1344
1345        if start in self._QUOTES:
1346            end = self._QUOTES[start]
1347        elif start in self._FORMAT_STRINGS:
1348            end, token_type = self._FORMAT_STRINGS[start]
1349
1350            if token_type == TokenType.HEX_STRING:
1351                base = 16
1352            elif token_type == TokenType.BIT_STRING:
1353                base = 2
1354            elif token_type == TokenType.HEREDOC_STRING:
1355                self._advance()
1356
1357                if self._char == end:
1358                    tag = ""
1359                else:
1360                    tag = self._extract_string(
1361                        end,
1362                        raw_string=True,
1363                        raise_unmatched=not self.HEREDOC_TAG_IS_IDENTIFIER,
1364                    )
1365
1366                if tag and self.HEREDOC_TAG_IS_IDENTIFIER and (self._end or not tag.isidentifier()):
1367                    if not self._end:
1368                        self._advance(-1)
1369
1370                    self._advance(-len(tag))
1371                    self._add(self.HEREDOC_STRING_ALTERNATIVE)
1372                    return True
1373
1374                end = f"{start}{tag}{end}"
1375        else:
1376            return False
1377
1378        self._advance(len(start))
1379        text = self._extract_string(end, raw_string=token_type == TokenType.RAW_STRING)
1380
1381        if base:
1382            try:
1383                int(text, base)
1384            except Exception:
1385                raise TokenError(
1386                    f"Numeric string contains invalid characters from {self._line}:{self._start}"
1387                )
1388
1389        self._add(token_type, text)
1390        return True
1391
1392    def _scan_identifier(self, identifier_end: str) -> None:
1393        self._advance()
1394        text = self._extract_string(
1395            identifier_end, escapes=self._IDENTIFIER_ESCAPES | {identifier_end}
1396        )
1397        self._add(TokenType.IDENTIFIER, text)
1398
1399    def _scan_var(self) -> None:
1400        while True:
1401            char = self._peek.strip()
1402            if char and (char in self.VAR_SINGLE_TOKENS or char not in self.SINGLE_TOKENS):
1403                self._advance(alnum=True)
1404            else:
1405                break
1406
1407        self._add(
1408            TokenType.VAR
1409            if self.tokens and self.tokens[-1].token_type == TokenType.PARAMETER
1410            else self.KEYWORDS.get(self._text.upper(), TokenType.VAR)
1411        )
1412
1413    def _extract_string(
1414        self,
1415        delimiter: str,
1416        escapes: t.Optional[t.Set[str]] = None,
1417        raw_string: bool = False,
1418        raise_unmatched: bool = True,
1419    ) -> str:
1420        text = ""
1421        delim_size = len(delimiter)
1422        escapes = self._STRING_ESCAPES if escapes is None else escapes
1423
1424        while True:
1425            if (
1426                not raw_string
1427                and self.dialect.UNESCAPED_SEQUENCES
1428                and self._peek
1429                and self._char in self.STRING_ESCAPES
1430            ):
1431                unescaped_sequence = self.dialect.UNESCAPED_SEQUENCES.get(self._char + self._peek)
1432                if unescaped_sequence:
1433                    self._advance(2)
1434                    text += unescaped_sequence
1435                    continue
1436            if (
1437                (self.STRING_ESCAPES_ALLOWED_IN_RAW_STRINGS or not raw_string)
1438                and self._char in escapes
1439                and (self._peek == delimiter or self._peek in escapes)
1440                and (self._char not in self._QUOTES or self._char == self._peek)
1441            ):
1442                if self._peek == delimiter:
1443                    text += self._peek
1444                else:
1445                    text += self._char + self._peek
1446
1447                if self._current + 1 < self.size:
1448                    self._advance(2)
1449                else:
1450                    raise TokenError(f"Missing {delimiter} from {self._line}:{self._current}")
1451            else:
1452                if self._chars(delim_size) == delimiter:
1453                    if delim_size > 1:
1454                        self._advance(delim_size - 1)
1455                    break
1456
1457                if self._end:
1458                    if not raise_unmatched:
1459                        return text + self._char
1460
1461                    raise TokenError(f"Missing {delimiter} from {self._line}:{self._start}")
1462
1463                current = self._current - 1
1464                self._advance(alnum=True)
1465                text += self.sql[current : self._current - 1]
1466
1467        return text
1468
1469    def tokenize_rs(self, sql: str) -> t.List[Token]:
1470        if not self._RS_TOKENIZER:
1471            raise SqlglotError("Rust tokenizer is not available")
1472
1473        try:
1474            tokens = self._RS_TOKENIZER.tokenize(sql, self._rs_dialect_settings)
1475            for token in tokens:
1476                token.token_type = _ALL_TOKEN_TYPES[token.token_type_index]
1477            return tokens
1478        except Exception as e:
1479            raise TokenError(str(e))