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