sqlglot.tokens

   1from __future__ import annotations
   2
   3import typing as t
   4from enum import auto
   5
   6from sqlglot.helper import AutoName
   7from sqlglot.trie import TrieResult, in_trie, new_trie
   8
   9
  10class TokenType(AutoName):
  11    L_PAREN = auto()
  12    R_PAREN = auto()
  13    L_BRACKET = auto()
  14    R_BRACKET = auto()
  15    L_BRACE = auto()
  16    R_BRACE = auto()
  17    COMMA = auto()
  18    DOT = auto()
  19    DASH = auto()
  20    PLUS = auto()
  21    COLON = auto()
  22    DCOLON = auto()
  23    SEMICOLON = auto()
  24    STAR = auto()
  25    BACKSLASH = auto()
  26    SLASH = auto()
  27    LT = auto()
  28    LTE = auto()
  29    GT = auto()
  30    GTE = auto()
  31    NOT = auto()
  32    EQ = auto()
  33    NEQ = auto()
  34    NULLSAFE_EQ = auto()
  35    AND = auto()
  36    OR = auto()
  37    AMP = auto()
  38    DPIPE = auto()
  39    PIPE = auto()
  40    CARET = auto()
  41    TILDA = auto()
  42    ARROW = auto()
  43    DARROW = auto()
  44    FARROW = auto()
  45    HASH = auto()
  46    HASH_ARROW = auto()
  47    DHASH_ARROW = auto()
  48    LR_ARROW = auto()
  49    LT_AT = auto()
  50    AT_GT = auto()
  51    DOLLAR = auto()
  52    PARAMETER = auto()
  53    SESSION_PARAMETER = auto()
  54    DAMP = auto()
  55    XOR = auto()
  56
  57    BLOCK_START = auto()
  58    BLOCK_END = auto()
  59
  60    SPACE = auto()
  61    BREAK = auto()
  62
  63    STRING = auto()
  64    NUMBER = auto()
  65    IDENTIFIER = auto()
  66    DATABASE = auto()
  67    COLUMN = auto()
  68    COLUMN_DEF = auto()
  69    SCHEMA = auto()
  70    TABLE = auto()
  71    VAR = auto()
  72    BIT_STRING = auto()
  73    HEX_STRING = auto()
  74    BYTE_STRING = auto()
  75    NATIONAL_STRING = auto()
  76    RAW_STRING = auto()
  77
  78    # types
  79    BIT = auto()
  80    BOOLEAN = auto()
  81    TINYINT = auto()
  82    UTINYINT = auto()
  83    SMALLINT = auto()
  84    USMALLINT = auto()
  85    INT = auto()
  86    UINT = auto()
  87    BIGINT = auto()
  88    UBIGINT = auto()
  89    INT128 = auto()
  90    UINT128 = auto()
  91    INT256 = auto()
  92    UINT256 = auto()
  93    FLOAT = auto()
  94    DOUBLE = auto()
  95    DECIMAL = auto()
  96    BIGDECIMAL = auto()
  97    CHAR = auto()
  98    NCHAR = auto()
  99    VARCHAR = auto()
 100    NVARCHAR = auto()
 101    TEXT = auto()
 102    MEDIUMTEXT = auto()
 103    LONGTEXT = auto()
 104    MEDIUMBLOB = auto()
 105    LONGBLOB = auto()
 106    BINARY = auto()
 107    VARBINARY = auto()
 108    JSON = auto()
 109    JSONB = auto()
 110    TIME = auto()
 111    TIMESTAMP = auto()
 112    TIMESTAMPTZ = auto()
 113    TIMESTAMPLTZ = auto()
 114    DATETIME = auto()
 115    DATETIME64 = auto()
 116    DATE = auto()
 117    INT4RANGE = auto()
 118    INT4MULTIRANGE = auto()
 119    INT8RANGE = auto()
 120    INT8MULTIRANGE = auto()
 121    NUMRANGE = auto()
 122    NUMMULTIRANGE = auto()
 123    TSRANGE = auto()
 124    TSMULTIRANGE = auto()
 125    TSTZRANGE = auto()
 126    TSTZMULTIRANGE = auto()
 127    DATERANGE = auto()
 128    DATEMULTIRANGE = auto()
 129    UUID = auto()
 130    GEOGRAPHY = auto()
 131    NULLABLE = auto()
 132    GEOMETRY = auto()
 133    HLLSKETCH = auto()
 134    HSTORE = auto()
 135    SUPER = auto()
 136    SERIAL = auto()
 137    SMALLSERIAL = auto()
 138    BIGSERIAL = auto()
 139    XML = auto()
 140    UNIQUEIDENTIFIER = auto()
 141    USERDEFINED = auto()
 142    MONEY = auto()
 143    SMALLMONEY = auto()
 144    ROWVERSION = auto()
 145    IMAGE = auto()
 146    VARIANT = auto()
 147    OBJECT = auto()
 148    INET = auto()
 149    ENUM = auto()
 150
 151    # keywords
 152    ALIAS = auto()
 153    ALTER = auto()
 154    ALWAYS = auto()
 155    ALL = auto()
 156    ANTI = auto()
 157    ANY = auto()
 158    APPLY = auto()
 159    ARRAY = auto()
 160    ASC = auto()
 161    ASOF = auto()
 162    AUTO_INCREMENT = auto()
 163    BEGIN = auto()
 164    BETWEEN = auto()
 165    CACHE = auto()
 166    CASE = auto()
 167    CHARACTER_SET = auto()
 168    CLUSTER_BY = auto()
 169    COLLATE = auto()
 170    COMMAND = auto()
 171    COMMENT = auto()
 172    COMMIT = auto()
 173    CONSTRAINT = auto()
 174    CREATE = auto()
 175    CROSS = auto()
 176    CUBE = auto()
 177    CURRENT_DATE = auto()
 178    CURRENT_DATETIME = auto()
 179    CURRENT_TIME = auto()
 180    CURRENT_TIMESTAMP = auto()
 181    CURRENT_USER = auto()
 182    DEFAULT = auto()
 183    DELETE = auto()
 184    DESC = auto()
 185    DESCRIBE = auto()
 186    DICTIONARY = auto()
 187    DISTINCT = auto()
 188    DISTRIBUTE_BY = auto()
 189    DIV = auto()
 190    DROP = auto()
 191    ELSE = auto()
 192    END = auto()
 193    ESCAPE = auto()
 194    EXCEPT = auto()
 195    EXECUTE = auto()
 196    EXISTS = auto()
 197    FALSE = auto()
 198    FETCH = auto()
 199    FILTER = auto()
 200    FINAL = auto()
 201    FIRST = auto()
 202    FOR = auto()
 203    FORCE = auto()
 204    FOREIGN_KEY = auto()
 205    FORMAT = auto()
 206    FROM = auto()
 207    FULL = auto()
 208    FUNCTION = auto()
 209    GLOB = auto()
 210    GLOBAL = auto()
 211    GROUP_BY = auto()
 212    GROUPING_SETS = auto()
 213    HAVING = auto()
 214    HINT = auto()
 215    IF = auto()
 216    IGNORE = auto()
 217    ILIKE = auto()
 218    ILIKE_ANY = auto()
 219    IN = auto()
 220    INDEX = auto()
 221    INNER = auto()
 222    INSERT = auto()
 223    INTERSECT = auto()
 224    INTERVAL = auto()
 225    INTO = auto()
 226    INTRODUCER = auto()
 227    IRLIKE = auto()
 228    IS = auto()
 229    ISNULL = auto()
 230    JOIN = auto()
 231    JOIN_MARKER = auto()
 232    KEEP = auto()
 233    LANGUAGE = auto()
 234    LATERAL = auto()
 235    LEFT = auto()
 236    LIKE = auto()
 237    LIKE_ANY = auto()
 238    LIMIT = auto()
 239    LOAD = auto()
 240    LOCK = auto()
 241    MAP = auto()
 242    MATCH_RECOGNIZE = auto()
 243    MEMBER_OF = auto()
 244    MERGE = auto()
 245    MOD = auto()
 246    NATURAL = auto()
 247    NEXT = auto()
 248    NEXT_VALUE_FOR = auto()
 249    NOTNULL = auto()
 250    NULL = auto()
 251    OFFSET = auto()
 252    ON = auto()
 253    ORDER_BY = auto()
 254    ORDERED = auto()
 255    ORDINALITY = auto()
 256    OUTER = auto()
 257    OVER = auto()
 258    OVERLAPS = auto()
 259    OVERWRITE = auto()
 260    PARTITION = auto()
 261    PARTITION_BY = auto()
 262    PERCENT = auto()
 263    PIVOT = auto()
 264    PLACEHOLDER = auto()
 265    PRAGMA = auto()
 266    PRIMARY_KEY = auto()
 267    PROCEDURE = auto()
 268    PROPERTIES = auto()
 269    PSEUDO_TYPE = auto()
 270    QUALIFY = auto()
 271    QUOTE = auto()
 272    RANGE = auto()
 273    RECURSIVE = auto()
 274    REPLACE = auto()
 275    RETURNING = auto()
 276    REFERENCES = auto()
 277    RIGHT = auto()
 278    RLIKE = auto()
 279    ROLLBACK = auto()
 280    ROLLUP = auto()
 281    ROW = auto()
 282    ROWS = auto()
 283    SELECT = auto()
 284    SEMI = auto()
 285    SEPARATOR = auto()
 286    SERDE_PROPERTIES = auto()
 287    SET = auto()
 288    SETTINGS = auto()
 289    SHOW = auto()
 290    SIMILAR_TO = auto()
 291    SOME = auto()
 292    SORT_BY = auto()
 293    STRUCT = auto()
 294    TABLE_SAMPLE = auto()
 295    TEMPORARY = auto()
 296    TOP = auto()
 297    THEN = auto()
 298    TRUE = auto()
 299    UNCACHE = auto()
 300    UNION = auto()
 301    UNNEST = auto()
 302    UNPIVOT = auto()
 303    UPDATE = auto()
 304    USE = auto()
 305    USING = auto()
 306    VALUES = auto()
 307    VIEW = auto()
 308    VOLATILE = auto()
 309    WHEN = auto()
 310    WHERE = auto()
 311    WINDOW = auto()
 312    WITH = auto()
 313    UNIQUE = auto()
 314
 315
 316class Token:
 317    __slots__ = ("token_type", "text", "line", "col", "start", "end", "comments")
 318
 319    @classmethod
 320    def number(cls, number: int) -> Token:
 321        """Returns a NUMBER token with `number` as its text."""
 322        return cls(TokenType.NUMBER, str(number))
 323
 324    @classmethod
 325    def string(cls, string: str) -> Token:
 326        """Returns a STRING token with `string` as its text."""
 327        return cls(TokenType.STRING, string)
 328
 329    @classmethod
 330    def identifier(cls, identifier: str) -> Token:
 331        """Returns an IDENTIFIER token with `identifier` as its text."""
 332        return cls(TokenType.IDENTIFIER, identifier)
 333
 334    @classmethod
 335    def var(cls, var: str) -> Token:
 336        """Returns an VAR token with `var` as its text."""
 337        return cls(TokenType.VAR, var)
 338
 339    def __init__(
 340        self,
 341        token_type: TokenType,
 342        text: str,
 343        line: int = 1,
 344        col: int = 1,
 345        start: int = 0,
 346        end: int = 0,
 347        comments: t.List[str] = [],
 348    ) -> None:
 349        """Token initializer.
 350
 351        Args:
 352            token_type: The TokenType Enum.
 353            text: The text of the token.
 354            line: The line that the token ends on.
 355            col: The column that the token ends on.
 356            start: The start index of the token.
 357            end: The ending index of the token.
 358            comments: The comments to attach to the token.
 359        """
 360        self.token_type = token_type
 361        self.text = text
 362        self.line = line
 363        self.col = col
 364        self.start = start
 365        self.end = end
 366        self.comments = comments
 367
 368    def __repr__(self) -> str:
 369        attributes = ", ".join(f"{k}: {getattr(self, k)}" for k in self.__slots__)
 370        return f"<Token {attributes}>"
 371
 372
 373class _Tokenizer(type):
 374    def __new__(cls, clsname, bases, attrs):
 375        klass = super().__new__(cls, clsname, bases, attrs)
 376
 377        def _convert_quotes(arr: t.List[str | t.Tuple[str, str]]) -> t.Dict[str, str]:
 378            return dict(
 379                (item, item) if isinstance(item, str) else (item[0], item[1]) for item in arr
 380            )
 381
 382        def _quotes_to_format(
 383            token_type: TokenType, arr: t.List[str | t.Tuple[str, str]]
 384        ) -> t.Dict[str, t.Tuple[str, TokenType]]:
 385            return {k: (v, token_type) for k, v in _convert_quotes(arr).items()}
 386
 387        klass._QUOTES = _convert_quotes(klass.QUOTES)
 388        klass._IDENTIFIERS = _convert_quotes(klass.IDENTIFIERS)
 389
 390        klass._FORMAT_STRINGS = {
 391            **{
 392                p + s: (e, TokenType.NATIONAL_STRING)
 393                for s, e in klass._QUOTES.items()
 394                for p in ("n", "N")
 395            },
 396            **_quotes_to_format(TokenType.BIT_STRING, klass.BIT_STRINGS),
 397            **_quotes_to_format(TokenType.BYTE_STRING, klass.BYTE_STRINGS),
 398            **_quotes_to_format(TokenType.HEX_STRING, klass.HEX_STRINGS),
 399            **_quotes_to_format(TokenType.RAW_STRING, klass.RAW_STRINGS),
 400        }
 401
 402        klass._STRING_ESCAPES = set(klass.STRING_ESCAPES)
 403        klass._IDENTIFIER_ESCAPES = set(klass.IDENTIFIER_ESCAPES)
 404        klass._COMMENTS = {
 405            **dict(
 406                (comment, None) if isinstance(comment, str) else (comment[0], comment[1])
 407                for comment in klass.COMMENTS
 408            ),
 409            "{#": "#}",  # Ensure Jinja comments are tokenized correctly in all dialects
 410        }
 411
 412        klass._KEYWORD_TRIE = new_trie(
 413            key.upper()
 414            for key in (
 415                *klass.KEYWORDS,
 416                *klass._COMMENTS,
 417                *klass._QUOTES,
 418                *klass._FORMAT_STRINGS,
 419            )
 420            if " " in key or any(single in key for single in klass.SINGLE_TOKENS)
 421        )
 422
 423        return klass
 424
 425
 426class Tokenizer(metaclass=_Tokenizer):
 427    SINGLE_TOKENS = {
 428        "(": TokenType.L_PAREN,
 429        ")": TokenType.R_PAREN,
 430        "[": TokenType.L_BRACKET,
 431        "]": TokenType.R_BRACKET,
 432        "{": TokenType.L_BRACE,
 433        "}": TokenType.R_BRACE,
 434        "&": TokenType.AMP,
 435        "^": TokenType.CARET,
 436        ":": TokenType.COLON,
 437        ",": TokenType.COMMA,
 438        ".": TokenType.DOT,
 439        "-": TokenType.DASH,
 440        "=": TokenType.EQ,
 441        ">": TokenType.GT,
 442        "<": TokenType.LT,
 443        "%": TokenType.MOD,
 444        "!": TokenType.NOT,
 445        "|": TokenType.PIPE,
 446        "+": TokenType.PLUS,
 447        ";": TokenType.SEMICOLON,
 448        "/": TokenType.SLASH,
 449        "\\": TokenType.BACKSLASH,
 450        "*": TokenType.STAR,
 451        "~": TokenType.TILDA,
 452        "?": TokenType.PLACEHOLDER,
 453        "@": TokenType.PARAMETER,
 454        # used for breaking a var like x'y' but nothing else
 455        # the token type doesn't matter
 456        "'": TokenType.QUOTE,
 457        "`": TokenType.IDENTIFIER,
 458        '"': TokenType.IDENTIFIER,
 459        "#": TokenType.HASH,
 460    }
 461
 462    BIT_STRINGS: t.List[str | t.Tuple[str, str]] = []
 463    BYTE_STRINGS: t.List[str | t.Tuple[str, str]] = []
 464    HEX_STRINGS: t.List[str | t.Tuple[str, str]] = []
 465    RAW_STRINGS: t.List[str | t.Tuple[str, str]] = []
 466    IDENTIFIERS: t.List[str | t.Tuple[str, str]] = ['"']
 467    IDENTIFIER_ESCAPES = ['"']
 468    QUOTES: t.List[t.Tuple[str, str] | str] = ["'"]
 469    STRING_ESCAPES = ["'"]
 470    VAR_SINGLE_TOKENS: t.Set[str] = set()
 471
 472    # Autofilled
 473    IDENTIFIERS_CAN_START_WITH_DIGIT: bool = False
 474
 475    _COMMENTS: t.Dict[str, str] = {}
 476    _FORMAT_STRINGS: t.Dict[str, t.Tuple[str, TokenType]] = {}
 477    _IDENTIFIERS: t.Dict[str, str] = {}
 478    _IDENTIFIER_ESCAPES: t.Set[str] = set()
 479    _QUOTES: t.Dict[str, str] = {}
 480    _STRING_ESCAPES: t.Set[str] = set()
 481    _KEYWORD_TRIE: t.Dict = {}
 482
 483    KEYWORDS: t.Dict[str, TokenType] = {
 484        **{f"{{%{postfix}": TokenType.BLOCK_START for postfix in ("", "+", "-")},
 485        **{f"{prefix}%}}": TokenType.BLOCK_END for prefix in ("", "+", "-")},
 486        **{f"{{{{{postfix}": TokenType.BLOCK_START for postfix in ("+", "-")},
 487        **{f"{prefix}}}}}": TokenType.BLOCK_END for prefix in ("+", "-")},
 488        "/*+": TokenType.HINT,
 489        "==": TokenType.EQ,
 490        "::": TokenType.DCOLON,
 491        "||": TokenType.DPIPE,
 492        ">=": TokenType.GTE,
 493        "<=": TokenType.LTE,
 494        "<>": TokenType.NEQ,
 495        "!=": TokenType.NEQ,
 496        "<=>": TokenType.NULLSAFE_EQ,
 497        "->": TokenType.ARROW,
 498        "->>": TokenType.DARROW,
 499        "=>": TokenType.FARROW,
 500        "#>": TokenType.HASH_ARROW,
 501        "#>>": TokenType.DHASH_ARROW,
 502        "<->": TokenType.LR_ARROW,
 503        "&&": TokenType.DAMP,
 504        "ALL": TokenType.ALL,
 505        "ALWAYS": TokenType.ALWAYS,
 506        "AND": TokenType.AND,
 507        "ANTI": TokenType.ANTI,
 508        "ANY": TokenType.ANY,
 509        "ASC": TokenType.ASC,
 510        "AS": TokenType.ALIAS,
 511        "ASOF": TokenType.ASOF,
 512        "AUTOINCREMENT": TokenType.AUTO_INCREMENT,
 513        "AUTO_INCREMENT": TokenType.AUTO_INCREMENT,
 514        "BEGIN": TokenType.BEGIN,
 515        "BETWEEN": TokenType.BETWEEN,
 516        "CACHE": TokenType.CACHE,
 517        "UNCACHE": TokenType.UNCACHE,
 518        "CASE": TokenType.CASE,
 519        "CHARACTER SET": TokenType.CHARACTER_SET,
 520        "CLUSTER BY": TokenType.CLUSTER_BY,
 521        "COLLATE": TokenType.COLLATE,
 522        "COLUMN": TokenType.COLUMN,
 523        "COMMIT": TokenType.COMMIT,
 524        "CONSTRAINT": TokenType.CONSTRAINT,
 525        "CREATE": TokenType.CREATE,
 526        "CROSS": TokenType.CROSS,
 527        "CUBE": TokenType.CUBE,
 528        "CURRENT_DATE": TokenType.CURRENT_DATE,
 529        "CURRENT_TIME": TokenType.CURRENT_TIME,
 530        "CURRENT_TIMESTAMP": TokenType.CURRENT_TIMESTAMP,
 531        "CURRENT_USER": TokenType.CURRENT_USER,
 532        "DATABASE": TokenType.DATABASE,
 533        "DEFAULT": TokenType.DEFAULT,
 534        "DELETE": TokenType.DELETE,
 535        "DESC": TokenType.DESC,
 536        "DESCRIBE": TokenType.DESCRIBE,
 537        "DISTINCT": TokenType.DISTINCT,
 538        "DISTRIBUTE BY": TokenType.DISTRIBUTE_BY,
 539        "DIV": TokenType.DIV,
 540        "DROP": TokenType.DROP,
 541        "ELSE": TokenType.ELSE,
 542        "END": TokenType.END,
 543        "ESCAPE": TokenType.ESCAPE,
 544        "EXCEPT": TokenType.EXCEPT,
 545        "EXECUTE": TokenType.EXECUTE,
 546        "EXISTS": TokenType.EXISTS,
 547        "FALSE": TokenType.FALSE,
 548        "FETCH": TokenType.FETCH,
 549        "FILTER": TokenType.FILTER,
 550        "FIRST": TokenType.FIRST,
 551        "FULL": TokenType.FULL,
 552        "FUNCTION": TokenType.FUNCTION,
 553        "FOR": TokenType.FOR,
 554        "FOREIGN KEY": TokenType.FOREIGN_KEY,
 555        "FORMAT": TokenType.FORMAT,
 556        "FROM": TokenType.FROM,
 557        "GEOGRAPHY": TokenType.GEOGRAPHY,
 558        "GEOMETRY": TokenType.GEOMETRY,
 559        "GLOB": TokenType.GLOB,
 560        "GROUP BY": TokenType.GROUP_BY,
 561        "GROUPING SETS": TokenType.GROUPING_SETS,
 562        "HAVING": TokenType.HAVING,
 563        "IF": TokenType.IF,
 564        "ILIKE": TokenType.ILIKE,
 565        "IN": TokenType.IN,
 566        "INDEX": TokenType.INDEX,
 567        "INET": TokenType.INET,
 568        "INNER": TokenType.INNER,
 569        "INSERT": TokenType.INSERT,
 570        "INTERVAL": TokenType.INTERVAL,
 571        "INTERSECT": TokenType.INTERSECT,
 572        "INTO": TokenType.INTO,
 573        "IS": TokenType.IS,
 574        "ISNULL": TokenType.ISNULL,
 575        "JOIN": TokenType.JOIN,
 576        "KEEP": TokenType.KEEP,
 577        "LATERAL": TokenType.LATERAL,
 578        "LEFT": TokenType.LEFT,
 579        "LIKE": TokenType.LIKE,
 580        "LIMIT": TokenType.LIMIT,
 581        "LOAD": TokenType.LOAD,
 582        "LOCK": TokenType.LOCK,
 583        "MERGE": TokenType.MERGE,
 584        "NATURAL": TokenType.NATURAL,
 585        "NEXT": TokenType.NEXT,
 586        "NEXT VALUE FOR": TokenType.NEXT_VALUE_FOR,
 587        "NOT": TokenType.NOT,
 588        "NOTNULL": TokenType.NOTNULL,
 589        "NULL": TokenType.NULL,
 590        "OBJECT": TokenType.OBJECT,
 591        "OFFSET": TokenType.OFFSET,
 592        "ON": TokenType.ON,
 593        "OR": TokenType.OR,
 594        "XOR": TokenType.XOR,
 595        "ORDER BY": TokenType.ORDER_BY,
 596        "ORDINALITY": TokenType.ORDINALITY,
 597        "OUTER": TokenType.OUTER,
 598        "OVER": TokenType.OVER,
 599        "OVERLAPS": TokenType.OVERLAPS,
 600        "OVERWRITE": TokenType.OVERWRITE,
 601        "PARTITION": TokenType.PARTITION,
 602        "PARTITION BY": TokenType.PARTITION_BY,
 603        "PARTITIONED BY": TokenType.PARTITION_BY,
 604        "PARTITIONED_BY": TokenType.PARTITION_BY,
 605        "PERCENT": TokenType.PERCENT,
 606        "PIVOT": TokenType.PIVOT,
 607        "PRAGMA": TokenType.PRAGMA,
 608        "PRIMARY KEY": TokenType.PRIMARY_KEY,
 609        "PROCEDURE": TokenType.PROCEDURE,
 610        "QUALIFY": TokenType.QUALIFY,
 611        "RANGE": TokenType.RANGE,
 612        "RECURSIVE": TokenType.RECURSIVE,
 613        "REGEXP": TokenType.RLIKE,
 614        "REPLACE": TokenType.REPLACE,
 615        "RETURNING": TokenType.RETURNING,
 616        "REFERENCES": TokenType.REFERENCES,
 617        "RIGHT": TokenType.RIGHT,
 618        "RLIKE": TokenType.RLIKE,
 619        "ROLLBACK": TokenType.ROLLBACK,
 620        "ROLLUP": TokenType.ROLLUP,
 621        "ROW": TokenType.ROW,
 622        "ROWS": TokenType.ROWS,
 623        "SCHEMA": TokenType.SCHEMA,
 624        "SELECT": TokenType.SELECT,
 625        "SEMI": TokenType.SEMI,
 626        "SET": TokenType.SET,
 627        "SETTINGS": TokenType.SETTINGS,
 628        "SHOW": TokenType.SHOW,
 629        "SIMILAR TO": TokenType.SIMILAR_TO,
 630        "SOME": TokenType.SOME,
 631        "SORT BY": TokenType.SORT_BY,
 632        "TABLE": TokenType.TABLE,
 633        "TABLESAMPLE": TokenType.TABLE_SAMPLE,
 634        "TEMP": TokenType.TEMPORARY,
 635        "TEMPORARY": TokenType.TEMPORARY,
 636        "THEN": TokenType.THEN,
 637        "TRUE": TokenType.TRUE,
 638        "UNION": TokenType.UNION,
 639        "UNNEST": TokenType.UNNEST,
 640        "UNPIVOT": TokenType.UNPIVOT,
 641        "UPDATE": TokenType.UPDATE,
 642        "USE": TokenType.USE,
 643        "USING": TokenType.USING,
 644        "UUID": TokenType.UUID,
 645        "VALUES": TokenType.VALUES,
 646        "VIEW": TokenType.VIEW,
 647        "VOLATILE": TokenType.VOLATILE,
 648        "WHEN": TokenType.WHEN,
 649        "WHERE": TokenType.WHERE,
 650        "WINDOW": TokenType.WINDOW,
 651        "WITH": TokenType.WITH,
 652        "APPLY": TokenType.APPLY,
 653        "ARRAY": TokenType.ARRAY,
 654        "BIT": TokenType.BIT,
 655        "BOOL": TokenType.BOOLEAN,
 656        "BOOLEAN": TokenType.BOOLEAN,
 657        "BYTE": TokenType.TINYINT,
 658        "TINYINT": TokenType.TINYINT,
 659        "SHORT": TokenType.SMALLINT,
 660        "SMALLINT": TokenType.SMALLINT,
 661        "INT2": TokenType.SMALLINT,
 662        "INTEGER": TokenType.INT,
 663        "INT": TokenType.INT,
 664        "INT4": TokenType.INT,
 665        "LONG": TokenType.BIGINT,
 666        "BIGINT": TokenType.BIGINT,
 667        "INT8": TokenType.BIGINT,
 668        "DEC": TokenType.DECIMAL,
 669        "DECIMAL": TokenType.DECIMAL,
 670        "BIGDECIMAL": TokenType.BIGDECIMAL,
 671        "BIGNUMERIC": TokenType.BIGDECIMAL,
 672        "MAP": TokenType.MAP,
 673        "NULLABLE": TokenType.NULLABLE,
 674        "NUMBER": TokenType.DECIMAL,
 675        "NUMERIC": TokenType.DECIMAL,
 676        "FIXED": TokenType.DECIMAL,
 677        "REAL": TokenType.FLOAT,
 678        "FLOAT": TokenType.FLOAT,
 679        "FLOAT4": TokenType.FLOAT,
 680        "FLOAT8": TokenType.DOUBLE,
 681        "DOUBLE": TokenType.DOUBLE,
 682        "DOUBLE PRECISION": TokenType.DOUBLE,
 683        "JSON": TokenType.JSON,
 684        "CHAR": TokenType.CHAR,
 685        "CHARACTER": TokenType.CHAR,
 686        "NCHAR": TokenType.NCHAR,
 687        "VARCHAR": TokenType.VARCHAR,
 688        "VARCHAR2": TokenType.VARCHAR,
 689        "NVARCHAR": TokenType.NVARCHAR,
 690        "NVARCHAR2": TokenType.NVARCHAR,
 691        "STR": TokenType.TEXT,
 692        "STRING": TokenType.TEXT,
 693        "TEXT": TokenType.TEXT,
 694        "CLOB": TokenType.TEXT,
 695        "LONGVARCHAR": TokenType.TEXT,
 696        "BINARY": TokenType.BINARY,
 697        "BLOB": TokenType.VARBINARY,
 698        "BYTEA": TokenType.VARBINARY,
 699        "VARBINARY": TokenType.VARBINARY,
 700        "TIME": TokenType.TIME,
 701        "TIMESTAMP": TokenType.TIMESTAMP,
 702        "TIMESTAMPTZ": TokenType.TIMESTAMPTZ,
 703        "TIMESTAMPLTZ": TokenType.TIMESTAMPLTZ,
 704        "DATE": TokenType.DATE,
 705        "DATETIME": TokenType.DATETIME,
 706        "INT4RANGE": TokenType.INT4RANGE,
 707        "INT4MULTIRANGE": TokenType.INT4MULTIRANGE,
 708        "INT8RANGE": TokenType.INT8RANGE,
 709        "INT8MULTIRANGE": TokenType.INT8MULTIRANGE,
 710        "NUMRANGE": TokenType.NUMRANGE,
 711        "NUMMULTIRANGE": TokenType.NUMMULTIRANGE,
 712        "TSRANGE": TokenType.TSRANGE,
 713        "TSMULTIRANGE": TokenType.TSMULTIRANGE,
 714        "TSTZRANGE": TokenType.TSTZRANGE,
 715        "TSTZMULTIRANGE": TokenType.TSTZMULTIRANGE,
 716        "DATERANGE": TokenType.DATERANGE,
 717        "DATEMULTIRANGE": TokenType.DATEMULTIRANGE,
 718        "UNIQUE": TokenType.UNIQUE,
 719        "STRUCT": TokenType.STRUCT,
 720        "VARIANT": TokenType.VARIANT,
 721        "ALTER": TokenType.ALTER,
 722        "ANALYZE": TokenType.COMMAND,
 723        "CALL": TokenType.COMMAND,
 724        "COMMENT": TokenType.COMMENT,
 725        "COPY": TokenType.COMMAND,
 726        "EXPLAIN": TokenType.COMMAND,
 727        "GRANT": TokenType.COMMAND,
 728        "OPTIMIZE": TokenType.COMMAND,
 729        "PREPARE": TokenType.COMMAND,
 730        "TRUNCATE": TokenType.COMMAND,
 731        "VACUUM": TokenType.COMMAND,
 732        "USER-DEFINED": TokenType.USERDEFINED,
 733    }
 734
 735    WHITE_SPACE: t.Dict[t.Optional[str], TokenType] = {
 736        " ": TokenType.SPACE,
 737        "\t": TokenType.SPACE,
 738        "\n": TokenType.BREAK,
 739        "\r": TokenType.BREAK,
 740        "\r\n": TokenType.BREAK,
 741    }
 742
 743    COMMANDS = {
 744        TokenType.COMMAND,
 745        TokenType.EXECUTE,
 746        TokenType.FETCH,
 747        TokenType.SHOW,
 748    }
 749
 750    COMMAND_PREFIX_TOKENS = {TokenType.SEMICOLON, TokenType.BEGIN}
 751
 752    # handle numeric literals like in hive (3L = BIGINT)
 753    NUMERIC_LITERALS: t.Dict[str, str] = {}
 754    ENCODE: t.Optional[str] = None
 755
 756    COMMENTS = ["--", ("/*", "*/")]
 757
 758    __slots__ = (
 759        "sql",
 760        "size",
 761        "tokens",
 762        "_start",
 763        "_current",
 764        "_line",
 765        "_col",
 766        "_comments",
 767        "_char",
 768        "_end",
 769        "_peek",
 770        "_prev_token_line",
 771    )
 772
 773    def __init__(self) -> None:
 774        self.reset()
 775
 776    def reset(self) -> None:
 777        self.sql = ""
 778        self.size = 0
 779        self.tokens: t.List[Token] = []
 780        self._start = 0
 781        self._current = 0
 782        self._line = 1
 783        self._col = 0
 784        self._comments: t.List[str] = []
 785
 786        self._char = ""
 787        self._end = False
 788        self._peek = ""
 789        self._prev_token_line = -1
 790
 791    def tokenize(self, sql: str) -> t.List[Token]:
 792        """Returns a list of tokens corresponding to the SQL string `sql`."""
 793        self.reset()
 794        self.sql = sql
 795        self.size = len(sql)
 796
 797        try:
 798            self._scan()
 799        except Exception as e:
 800            start = max(self._current - 50, 0)
 801            end = min(self._current + 50, self.size - 1)
 802            context = self.sql[start:end]
 803            raise ValueError(f"Error tokenizing '{context}'") from e
 804
 805        return self.tokens
 806
 807    def _scan(self, until: t.Optional[t.Callable] = None) -> None:
 808        while self.size and not self._end:
 809            self._start = self._current
 810            self._advance()
 811
 812            if self._char is None:
 813                break
 814
 815            if self._char not in self.WHITE_SPACE:
 816                if self._char.isdigit():
 817                    self._scan_number()
 818                elif self._char in self._IDENTIFIERS:
 819                    self._scan_identifier(self._IDENTIFIERS[self._char])
 820                else:
 821                    self._scan_keywords()
 822
 823            if until and until():
 824                break
 825
 826        if self.tokens and self._comments:
 827            self.tokens[-1].comments.extend(self._comments)
 828
 829    def _chars(self, size: int) -> str:
 830        if size == 1:
 831            return self._char
 832
 833        start = self._current - 1
 834        end = start + size
 835
 836        return self.sql[start:end] if end <= self.size else ""
 837
 838    def _advance(self, i: int = 1, alnum: bool = False) -> None:
 839        if self.WHITE_SPACE.get(self._char) is TokenType.BREAK:
 840            self._col = 1
 841            self._line += 1
 842        else:
 843            self._col += i
 844
 845        self._current += i
 846        self._end = self._current >= self.size
 847        self._char = self.sql[self._current - 1]
 848        self._peek = "" if self._end else self.sql[self._current]
 849
 850        if alnum and self._char.isalnum():
 851            # Here we use local variables instead of attributes for better performance
 852            _col = self._col
 853            _current = self._current
 854            _end = self._end
 855            _peek = self._peek
 856
 857            while _peek.isalnum():
 858                _col += 1
 859                _current += 1
 860                _end = _current >= self.size
 861                _peek = "" if _end else self.sql[_current]
 862
 863            self._col = _col
 864            self._current = _current
 865            self._end = _end
 866            self._peek = _peek
 867            self._char = self.sql[_current - 1]
 868
 869    @property
 870    def _text(self) -> str:
 871        return self.sql[self._start : self._current]
 872
 873    def peek(self, i: int = 0) -> str:
 874        i = self._current + i
 875        if i < self.size:
 876            return self.sql[i]
 877        return ""
 878
 879    def _add(self, token_type: TokenType, text: t.Optional[str] = None) -> None:
 880        self._prev_token_line = self._line
 881        self.tokens.append(
 882            Token(
 883                token_type,
 884                text=self._text if text is None else text,
 885                line=self._line,
 886                col=self._col,
 887                start=self._start,
 888                end=self._current - 1,
 889                comments=self._comments,
 890            )
 891        )
 892        self._comments = []
 893
 894        # If we have either a semicolon or a begin token before the command's token, we'll parse
 895        # whatever follows the command's token as a string
 896        if (
 897            token_type in self.COMMANDS
 898            and self._peek != ";"
 899            and (len(self.tokens) == 1 or self.tokens[-2].token_type in self.COMMAND_PREFIX_TOKENS)
 900        ):
 901            start = self._current
 902            tokens = len(self.tokens)
 903            self._scan(lambda: self._peek == ";")
 904            self.tokens = self.tokens[:tokens]
 905            text = self.sql[start : self._current].strip()
 906            if text:
 907                self._add(TokenType.STRING, text)
 908
 909    def _scan_keywords(self) -> None:
 910        size = 0
 911        word = None
 912        chars = self._text
 913        char = chars
 914        prev_space = False
 915        skip = False
 916        trie = self._KEYWORD_TRIE
 917        single_token = char in self.SINGLE_TOKENS
 918
 919        while chars:
 920            if skip:
 921                result = TrieResult.PREFIX
 922            else:
 923                result, trie = in_trie(trie, char.upper())
 924
 925            if result == TrieResult.FAILED:
 926                break
 927            if result == TrieResult.EXISTS:
 928                word = chars
 929
 930            size += 1
 931            end = self._current - 1 + size
 932
 933            if end < self.size:
 934                char = self.sql[end]
 935                single_token = single_token or char in self.SINGLE_TOKENS
 936                is_space = char in self.WHITE_SPACE
 937
 938                if not is_space or not prev_space:
 939                    if is_space:
 940                        char = " "
 941                    chars += char
 942                    prev_space = is_space
 943                    skip = False
 944                else:
 945                    skip = True
 946            else:
 947                char = ""
 948                chars = " "
 949
 950        if not word:
 951            if self._char in self.SINGLE_TOKENS:
 952                self._add(self.SINGLE_TOKENS[self._char], text=self._char)
 953                return
 954            self._scan_var()
 955            return
 956
 957        if self._scan_string(word):
 958            return
 959        if self._scan_comment(word):
 960            return
 961
 962        self._advance(size - 1)
 963        word = word.upper()
 964        self._add(self.KEYWORDS[word], text=word)
 965
 966    def _scan_comment(self, comment_start: str) -> bool:
 967        if comment_start not in self._COMMENTS:
 968            return False
 969
 970        comment_start_line = self._line
 971        comment_start_size = len(comment_start)
 972        comment_end = self._COMMENTS[comment_start]
 973
 974        if comment_end:
 975            # Skip the comment's start delimiter
 976            self._advance(comment_start_size)
 977
 978            comment_end_size = len(comment_end)
 979            while not self._end and self._chars(comment_end_size) != comment_end:
 980                self._advance(alnum=True)
 981
 982            self._comments.append(self._text[comment_start_size : -comment_end_size + 1])
 983            self._advance(comment_end_size - 1)
 984        else:
 985            while not self._end and not self.WHITE_SPACE.get(self._peek) is TokenType.BREAK:
 986                self._advance(alnum=True)
 987            self._comments.append(self._text[comment_start_size:])
 988
 989        # Leading comment is attached to the succeeding token, whilst trailing comment to the preceding.
 990        # Multiple consecutive comments are preserved by appending them to the current comments list.
 991        if comment_start_line == self._prev_token_line:
 992            self.tokens[-1].comments.extend(self._comments)
 993            self._comments = []
 994            self._prev_token_line = self._line
 995
 996        return True
 997
 998    def _scan_number(self) -> None:
 999        if self._char == "0":
1000            peek = self._peek.upper()
1001            if peek == "B":
1002                return self._scan_bits() if self.BIT_STRINGS else self._add(TokenType.NUMBER)
1003            elif peek == "X":
1004                return self._scan_hex() if self.HEX_STRINGS else self._add(TokenType.NUMBER)
1005
1006        decimal = False
1007        scientific = 0
1008
1009        while True:
1010            if self._peek.isdigit():
1011                self._advance()
1012            elif self._peek == "." and not decimal:
1013                after = self.peek(1)
1014                if after.isdigit() or not after.isalpha():
1015                    decimal = True
1016                    self._advance()
1017                else:
1018                    return self._add(TokenType.VAR)
1019            elif self._peek in ("-", "+") and scientific == 1:
1020                scientific += 1
1021                self._advance()
1022            elif self._peek.upper() == "E" and not scientific:
1023                scientific += 1
1024                self._advance()
1025            elif self._peek.isidentifier():
1026                number_text = self._text
1027                literal = ""
1028
1029                while self._peek.strip() and self._peek not in self.SINGLE_TOKENS:
1030                    literal += self._peek.upper()
1031                    self._advance()
1032
1033                token_type = self.KEYWORDS.get(self.NUMERIC_LITERALS.get(literal, ""))
1034
1035                if token_type:
1036                    self._add(TokenType.NUMBER, number_text)
1037                    self._add(TokenType.DCOLON, "::")
1038                    return self._add(token_type, literal)
1039                elif self.IDENTIFIERS_CAN_START_WITH_DIGIT:
1040                    return self._add(TokenType.VAR)
1041
1042                self._add(TokenType.NUMBER, number_text)
1043                return self._advance(-len(literal))
1044            else:
1045                return self._add(TokenType.NUMBER)
1046
1047    def _scan_bits(self) -> None:
1048        self._advance()
1049        value = self._extract_value()
1050        try:
1051            # If `value` can't be converted to a binary, fallback to tokenizing it as an identifier
1052            int(value, 2)
1053            self._add(TokenType.BIT_STRING, value[2:])  # Drop the 0b
1054        except ValueError:
1055            self._add(TokenType.IDENTIFIER)
1056
1057    def _scan_hex(self) -> None:
1058        self._advance()
1059        value = self._extract_value()
1060        try:
1061            # If `value` can't be converted to a hex, fallback to tokenizing it as an identifier
1062            int(value, 16)
1063            self._add(TokenType.HEX_STRING, value[2:])  # Drop the 0x
1064        except ValueError:
1065            self._add(TokenType.IDENTIFIER)
1066
1067    def _extract_value(self) -> str:
1068        while True:
1069            char = self._peek.strip()
1070            if char and char not in self.SINGLE_TOKENS:
1071                self._advance(alnum=True)
1072            else:
1073                break
1074
1075        return self._text
1076
1077    def _scan_string(self, start: str) -> bool:
1078        base = None
1079        token_type = TokenType.STRING
1080
1081        if start in self._QUOTES:
1082            end = self._QUOTES[start]
1083        elif start in self._FORMAT_STRINGS:
1084            end, token_type = self._FORMAT_STRINGS[start]
1085
1086            if token_type == TokenType.HEX_STRING:
1087                base = 16
1088            elif token_type == TokenType.BIT_STRING:
1089                base = 2
1090        else:
1091            return False
1092
1093        self._advance(len(start))
1094        text = self._extract_string(end)
1095
1096        if base:
1097            try:
1098                int(text, base)
1099            except:
1100                raise RuntimeError(
1101                    f"Numeric string contains invalid characters from {self._line}:{self._start}"
1102                )
1103        else:
1104            text = text.encode(self.ENCODE).decode(self.ENCODE) if self.ENCODE else text
1105
1106        self._add(token_type, text)
1107        return True
1108
1109    def _scan_identifier(self, identifier_end: str) -> None:
1110        self._advance()
1111        text = self._extract_string(identifier_end, self._IDENTIFIER_ESCAPES)
1112        self._add(TokenType.IDENTIFIER, text)
1113
1114    def _scan_var(self) -> None:
1115        while True:
1116            char = self._peek.strip()
1117            if char and (char in self.VAR_SINGLE_TOKENS or char not in self.SINGLE_TOKENS):
1118                self._advance(alnum=True)
1119            else:
1120                break
1121
1122        self._add(
1123            TokenType.VAR
1124            if self.tokens and self.tokens[-1].token_type == TokenType.PARAMETER
1125            else self.KEYWORDS.get(self._text.upper(), TokenType.VAR)
1126        )
1127
1128    def _extract_string(self, delimiter: str, escapes=None) -> str:
1129        text = ""
1130        delim_size = len(delimiter)
1131        escapes = self._STRING_ESCAPES if escapes is None else escapes
1132
1133        while True:
1134            if self._char in escapes and (self._peek == delimiter or self._peek in escapes):
1135                if self._peek == delimiter:
1136                    text += self._peek
1137                else:
1138                    text += self._char + self._peek
1139
1140                if self._current + 1 < self.size:
1141                    self._advance(2)
1142                else:
1143                    raise RuntimeError(f"Missing {delimiter} from {self._line}:{self._current}")
1144            else:
1145                if self._chars(delim_size) == delimiter:
1146                    if delim_size > 1:
1147                        self._advance(delim_size - 1)
1148                    break
1149
1150                if self._end:
1151                    raise RuntimeError(f"Missing {delimiter} from {self._line}:{self._start}")
1152
1153                current = self._current - 1
1154                self._advance(alnum=True)
1155                text += self.sql[current : self._current - 1]
1156
1157        return text