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    NATIONAL = auto()
  55    DAMP = 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
  76    # types
  77    BIT = auto()
  78    BOOLEAN = auto()
  79    TINYINT = auto()
  80    UTINYINT = auto()
  81    SMALLINT = auto()
  82    USMALLINT = auto()
  83    INT = auto()
  84    UINT = auto()
  85    BIGINT = auto()
  86    UBIGINT = auto()
  87    FLOAT = auto()
  88    DOUBLE = auto()
  89    DECIMAL = auto()
  90    BIGDECIMAL = auto()
  91    CHAR = auto()
  92    NCHAR = auto()
  93    VARCHAR = auto()
  94    NVARCHAR = auto()
  95    TEXT = auto()
  96    MEDIUMTEXT = auto()
  97    LONGTEXT = auto()
  98    MEDIUMBLOB = auto()
  99    LONGBLOB = auto()
 100    BINARY = auto()
 101    VARBINARY = auto()
 102    JSON = auto()
 103    JSONB = auto()
 104    TIME = auto()
 105    TIMESTAMP = auto()
 106    TIMESTAMPTZ = auto()
 107    TIMESTAMPLTZ = auto()
 108    DATETIME = auto()
 109    DATE = auto()
 110    UUID = auto()
 111    GEOGRAPHY = auto()
 112    NULLABLE = auto()
 113    GEOMETRY = auto()
 114    HLLSKETCH = auto()
 115    HSTORE = auto()
 116    SUPER = auto()
 117    SERIAL = auto()
 118    SMALLSERIAL = auto()
 119    BIGSERIAL = auto()
 120    XML = auto()
 121    UNIQUEIDENTIFIER = auto()
 122    MONEY = auto()
 123    SMALLMONEY = auto()
 124    ROWVERSION = auto()
 125    IMAGE = auto()
 126    VARIANT = auto()
 127    OBJECT = auto()
 128    INET = auto()
 129
 130    # keywords
 131    ALIAS = auto()
 132    ALTER = auto()
 133    ALWAYS = auto()
 134    ALL = auto()
 135    ANTI = auto()
 136    ANY = auto()
 137    APPLY = auto()
 138    ARRAY = auto()
 139    ASC = auto()
 140    ASOF = auto()
 141    AT_TIME_ZONE = auto()
 142    AUTO_INCREMENT = auto()
 143    BEGIN = auto()
 144    BETWEEN = auto()
 145    BOTH = auto()
 146    BUCKET = auto()
 147    BY_DEFAULT = auto()
 148    CACHE = auto()
 149    CASCADE = auto()
 150    CASE = auto()
 151    CHARACTER_SET = auto()
 152    CLUSTER_BY = auto()
 153    COLLATE = auto()
 154    COMMAND = auto()
 155    COMMENT = auto()
 156    COMMIT = auto()
 157    COMPOUND = auto()
 158    CONSTRAINT = auto()
 159    CREATE = auto()
 160    CROSS = auto()
 161    CUBE = auto()
 162    CURRENT_DATE = auto()
 163    CURRENT_DATETIME = auto()
 164    CURRENT_ROW = auto()
 165    CURRENT_TIME = auto()
 166    CURRENT_TIMESTAMP = auto()
 167    CURRENT_USER = auto()
 168    DEFAULT = auto()
 169    DELETE = auto()
 170    DESC = auto()
 171    DESCRIBE = auto()
 172    DISTINCT = auto()
 173    DISTINCT_FROM = auto()
 174    DISTRIBUTE_BY = auto()
 175    DIV = auto()
 176    DROP = auto()
 177    ELSE = auto()
 178    END = auto()
 179    ESCAPE = auto()
 180    EXCEPT = auto()
 181    EXECUTE = auto()
 182    EXISTS = auto()
 183    FALSE = auto()
 184    FETCH = auto()
 185    FILTER = auto()
 186    FINAL = auto()
 187    FIRST = auto()
 188    FOLLOWING = auto()
 189    FOR = auto()
 190    FOREIGN_KEY = auto()
 191    FORMAT = auto()
 192    FROM = auto()
 193    FULL = auto()
 194    FUNCTION = auto()
 195    GLOB = auto()
 196    GLOBAL = auto()
 197    GROUP_BY = auto()
 198    GROUPING_SETS = auto()
 199    HAVING = auto()
 200    HINT = auto()
 201    IF = auto()
 202    IGNORE_NULLS = auto()
 203    ILIKE = auto()
 204    ILIKE_ANY = auto()
 205    IN = auto()
 206    INDEX = auto()
 207    INNER = auto()
 208    INSERT = auto()
 209    INTERSECT = auto()
 210    INTERVAL = auto()
 211    INTO = auto()
 212    INTRODUCER = auto()
 213    IRLIKE = auto()
 214    IS = auto()
 215    ISNULL = auto()
 216    JOIN = auto()
 217    JOIN_MARKER = auto()
 218    KEEP = auto()
 219    LANGUAGE = auto()
 220    LATERAL = auto()
 221    LAZY = auto()
 222    LEADING = auto()
 223    LEFT = auto()
 224    LIKE = auto()
 225    LIKE_ANY = auto()
 226    LIMIT = auto()
 227    LOAD_DATA = auto()
 228    LOCAL = auto()
 229    MAP = auto()
 230    MATCH_RECOGNIZE = auto()
 231    MATERIALIZED = auto()
 232    MERGE = auto()
 233    MOD = auto()
 234    NATURAL = auto()
 235    NEXT = auto()
 236    NEXT_VALUE_FOR = auto()
 237    NO_ACTION = auto()
 238    NOTNULL = auto()
 239    NULL = auto()
 240    NULLS_FIRST = auto()
 241    NULLS_LAST = auto()
 242    OFFSET = auto()
 243    ON = auto()
 244    ONLY = auto()
 245    OPTIONS = auto()
 246    ORDER_BY = auto()
 247    ORDERED = auto()
 248    ORDINALITY = auto()
 249    OUTER = auto()
 250    OUT_OF = auto()
 251    OVER = auto()
 252    OVERLAPS = auto()
 253    OVERWRITE = auto()
 254    PARTITION = auto()
 255    PARTITION_BY = auto()
 256    PERCENT = auto()
 257    PIVOT = auto()
 258    PLACEHOLDER = auto()
 259    PRAGMA = auto()
 260    PRECEDING = auto()
 261    PRIMARY_KEY = auto()
 262    PROCEDURE = auto()
 263    PROPERTIES = auto()
 264    PSEUDO_TYPE = auto()
 265    QUALIFY = auto()
 266    QUOTE = auto()
 267    RANGE = auto()
 268    RECURSIVE = auto()
 269    REPLACE = auto()
 270    RESPECT_NULLS = auto()
 271    RETURNING = auto()
 272    REFERENCES = auto()
 273    RIGHT = auto()
 274    RLIKE = auto()
 275    ROLLBACK = auto()
 276    ROLLUP = auto()
 277    ROW = auto()
 278    ROWS = auto()
 279    SEED = auto()
 280    SELECT = auto()
 281    SEMI = auto()
 282    SEPARATOR = auto()
 283    SERDE_PROPERTIES = auto()
 284    SET = auto()
 285    SHOW = auto()
 286    SIMILAR_TO = auto()
 287    SOME = auto()
 288    SORTKEY = auto()
 289    SORT_BY = auto()
 290    STRUCT = auto()
 291    TABLE_SAMPLE = auto()
 292    TEMPORARY = auto()
 293    TOP = auto()
 294    THEN = auto()
 295    TRAILING = auto()
 296    TRUE = auto()
 297    UNBOUNDED = auto()
 298    UNCACHE = auto()
 299    UNION = auto()
 300    UNLOGGED = 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    WITH_TIME_ZONE = auto()
 314    WITH_LOCAL_TIME_ZONE = auto()
 315    WITHIN_GROUP = auto()
 316    WITHOUT_TIME_ZONE = auto()
 317    UNIQUE = auto()
 318
 319
 320class Token:
 321    __slots__ = ("token_type", "text", "line", "col", "end", "comments")
 322
 323    @classmethod
 324    def number(cls, number: int) -> Token:
 325        """Returns a NUMBER token with `number` as its text."""
 326        return cls(TokenType.NUMBER, str(number))
 327
 328    @classmethod
 329    def string(cls, string: str) -> Token:
 330        """Returns a STRING token with `string` as its text."""
 331        return cls(TokenType.STRING, string)
 332
 333    @classmethod
 334    def identifier(cls, identifier: str) -> Token:
 335        """Returns an IDENTIFIER token with `identifier` as its text."""
 336        return cls(TokenType.IDENTIFIER, identifier)
 337
 338    @classmethod
 339    def var(cls, var: str) -> Token:
 340        """Returns an VAR token with `var` as its text."""
 341        return cls(TokenType.VAR, var)
 342
 343    def __init__(
 344        self,
 345        token_type: TokenType,
 346        text: str,
 347        line: int = 1,
 348        col: int = 1,
 349        end: int = 0,
 350        comments: t.List[str] = [],
 351    ) -> None:
 352        self.token_type = token_type
 353        self.text = text
 354        self.line = line
 355        size = len(text)
 356        self.col = col
 357        self.end = end if end else size
 358        self.comments = comments
 359
 360    @property
 361    def start(self) -> int:
 362        """Returns the start of the token."""
 363        return self.end - len(self.text)
 364
 365    def __repr__(self) -> str:
 366        attributes = ", ".join(f"{k}: {getattr(self, k)}" for k in self.__slots__)
 367        return f"<Token {attributes}>"
 368
 369
 370class _Tokenizer(type):
 371    def __new__(cls, clsname, bases, attrs):
 372        klass = super().__new__(cls, clsname, bases, attrs)
 373
 374        klass._QUOTES = {
 375            f"{prefix}{s}": e
 376            for s, e in cls._delimeter_list_to_dict(klass.QUOTES).items()
 377            for prefix in (("",) if s[0].isalpha() else ("", "n", "N"))
 378        }
 379        klass._BIT_STRINGS = cls._delimeter_list_to_dict(klass.BIT_STRINGS)
 380        klass._HEX_STRINGS = cls._delimeter_list_to_dict(klass.HEX_STRINGS)
 381        klass._BYTE_STRINGS = cls._delimeter_list_to_dict(klass.BYTE_STRINGS)
 382        klass._IDENTIFIERS = cls._delimeter_list_to_dict(klass.IDENTIFIERS)
 383        klass._STRING_ESCAPES = set(klass.STRING_ESCAPES)
 384        klass._IDENTIFIER_ESCAPES = set(klass.IDENTIFIER_ESCAPES)
 385        klass._COMMENTS = dict(
 386            (comment, None) if isinstance(comment, str) else (comment[0], comment[1])
 387            for comment in klass.COMMENTS
 388        )
 389
 390        klass.KEYWORD_TRIE = new_trie(
 391            key.upper()
 392            for key in {
 393                **klass.KEYWORDS,
 394                **{comment: TokenType.COMMENT for comment in klass._COMMENTS},
 395                **{quote: TokenType.QUOTE for quote in klass._QUOTES},
 396                **{bit_string: TokenType.BIT_STRING for bit_string in klass._BIT_STRINGS},
 397                **{hex_string: TokenType.HEX_STRING for hex_string in klass._HEX_STRINGS},
 398                **{byte_string: TokenType.BYTE_STRING for byte_string in klass._BYTE_STRINGS},
 399            }
 400            if " " in key or any(single in key for single in klass.SINGLE_TOKENS)
 401        )
 402
 403        return klass
 404
 405    @staticmethod
 406    def _delimeter_list_to_dict(list: t.List[str | t.Tuple[str, str]]) -> t.Dict[str, str]:
 407        return dict((item, item) if isinstance(item, str) else (item[0], item[1]) for item in list)
 408
 409
 410class Tokenizer(metaclass=_Tokenizer):
 411    SINGLE_TOKENS = {
 412        "(": TokenType.L_PAREN,
 413        ")": TokenType.R_PAREN,
 414        "[": TokenType.L_BRACKET,
 415        "]": TokenType.R_BRACKET,
 416        "{": TokenType.L_BRACE,
 417        "}": TokenType.R_BRACE,
 418        "&": TokenType.AMP,
 419        "^": TokenType.CARET,
 420        ":": TokenType.COLON,
 421        ",": TokenType.COMMA,
 422        ".": TokenType.DOT,
 423        "-": TokenType.DASH,
 424        "=": TokenType.EQ,
 425        ">": TokenType.GT,
 426        "<": TokenType.LT,
 427        "%": TokenType.MOD,
 428        "!": TokenType.NOT,
 429        "|": TokenType.PIPE,
 430        "+": TokenType.PLUS,
 431        ";": TokenType.SEMICOLON,
 432        "/": TokenType.SLASH,
 433        "\\": TokenType.BACKSLASH,
 434        "*": TokenType.STAR,
 435        "~": TokenType.TILDA,
 436        "?": TokenType.PLACEHOLDER,
 437        "@": TokenType.PARAMETER,
 438        # used for breaking a var like x'y' but nothing else
 439        # the token type doesn't matter
 440        "'": TokenType.QUOTE,
 441        "`": TokenType.IDENTIFIER,
 442        '"': TokenType.IDENTIFIER,
 443        "#": TokenType.HASH,
 444    }
 445
 446    BIT_STRINGS: t.List[str | t.Tuple[str, str]] = []
 447    BYTE_STRINGS: t.List[str | t.Tuple[str, str]] = []
 448    HEX_STRINGS: t.List[str | t.Tuple[str, str]] = []
 449    IDENTIFIERS: t.List[str | t.Tuple[str, str]] = ['"']
 450    IDENTIFIER_ESCAPES = ['"']
 451    QUOTES: t.List[t.Tuple[str, str] | str] = ["'"]
 452    STRING_ESCAPES = ["'"]
 453    VAR_SINGLE_TOKENS: t.Set[str] = set()
 454
 455    _COMMENTS: t.Dict[str, str] = {}
 456    _BIT_STRINGS: t.Dict[str, str] = {}
 457    _BYTE_STRINGS: t.Dict[str, str] = {}
 458    _HEX_STRINGS: t.Dict[str, str] = {}
 459    _IDENTIFIERS: t.Dict[str, str] = {}
 460    _IDENTIFIER_ESCAPES: t.Set[str] = set()
 461    _QUOTES: t.Dict[str, str] = {}
 462    _STRING_ESCAPES: t.Set[str] = set()
 463
 464    KEYWORDS: t.Dict[t.Optional[str], TokenType] = {
 465        **{f"{{%{postfix}": TokenType.BLOCK_START for postfix in ("", "+", "-")},
 466        **{f"{prefix}%}}": TokenType.BLOCK_END for prefix in ("", "+", "-")},
 467        "{{+": TokenType.BLOCK_START,
 468        "{{-": TokenType.BLOCK_START,
 469        "+}}": TokenType.BLOCK_END,
 470        "-}}": TokenType.BLOCK_END,
 471        "/*+": TokenType.HINT,
 472        "==": TokenType.EQ,
 473        "::": TokenType.DCOLON,
 474        "||": TokenType.DPIPE,
 475        ">=": TokenType.GTE,
 476        "<=": TokenType.LTE,
 477        "<>": TokenType.NEQ,
 478        "!=": TokenType.NEQ,
 479        "<=>": TokenType.NULLSAFE_EQ,
 480        "->": TokenType.ARROW,
 481        "->>": TokenType.DARROW,
 482        "=>": TokenType.FARROW,
 483        "#>": TokenType.HASH_ARROW,
 484        "#>>": TokenType.DHASH_ARROW,
 485        "<->": TokenType.LR_ARROW,
 486        "&&": TokenType.DAMP,
 487        "ALL": TokenType.ALL,
 488        "ALWAYS": TokenType.ALWAYS,
 489        "AND": TokenType.AND,
 490        "ANTI": TokenType.ANTI,
 491        "ANY": TokenType.ANY,
 492        "ASC": TokenType.ASC,
 493        "AS": TokenType.ALIAS,
 494        "AT TIME ZONE": TokenType.AT_TIME_ZONE,
 495        "AUTOINCREMENT": TokenType.AUTO_INCREMENT,
 496        "AUTO_INCREMENT": TokenType.AUTO_INCREMENT,
 497        "BEGIN": TokenType.BEGIN,
 498        "BETWEEN": TokenType.BETWEEN,
 499        "BOTH": TokenType.BOTH,
 500        "BUCKET": TokenType.BUCKET,
 501        "BY DEFAULT": TokenType.BY_DEFAULT,
 502        "CACHE": TokenType.CACHE,
 503        "UNCACHE": TokenType.UNCACHE,
 504        "CASE": TokenType.CASE,
 505        "CASCADE": TokenType.CASCADE,
 506        "CHARACTER SET": TokenType.CHARACTER_SET,
 507        "CLUSTER BY": TokenType.CLUSTER_BY,
 508        "COLLATE": TokenType.COLLATE,
 509        "COLUMN": TokenType.COLUMN,
 510        "COMMIT": TokenType.COMMIT,
 511        "COMPOUND": TokenType.COMPOUND,
 512        "CONSTRAINT": TokenType.CONSTRAINT,
 513        "CREATE": TokenType.CREATE,
 514        "CROSS": TokenType.CROSS,
 515        "CUBE": TokenType.CUBE,
 516        "CURRENT_DATE": TokenType.CURRENT_DATE,
 517        "CURRENT ROW": TokenType.CURRENT_ROW,
 518        "CURRENT_TIME": TokenType.CURRENT_TIME,
 519        "CURRENT_TIMESTAMP": TokenType.CURRENT_TIMESTAMP,
 520        "CURRENT_USER": TokenType.CURRENT_USER,
 521        "DATABASE": TokenType.DATABASE,
 522        "DEFAULT": TokenType.DEFAULT,
 523        "DELETE": TokenType.DELETE,
 524        "DESC": TokenType.DESC,
 525        "DESCRIBE": TokenType.DESCRIBE,
 526        "DISTINCT": TokenType.DISTINCT,
 527        "DISTINCT FROM": TokenType.DISTINCT_FROM,
 528        "DISTRIBUTE BY": TokenType.DISTRIBUTE_BY,
 529        "DIV": TokenType.DIV,
 530        "DROP": TokenType.DROP,
 531        "ELSE": TokenType.ELSE,
 532        "END": TokenType.END,
 533        "ESCAPE": TokenType.ESCAPE,
 534        "EXCEPT": TokenType.EXCEPT,
 535        "EXECUTE": TokenType.EXECUTE,
 536        "EXISTS": TokenType.EXISTS,
 537        "FALSE": TokenType.FALSE,
 538        "FETCH": TokenType.FETCH,
 539        "FILTER": TokenType.FILTER,
 540        "FIRST": TokenType.FIRST,
 541        "FULL": TokenType.FULL,
 542        "FUNCTION": TokenType.FUNCTION,
 543        "FOLLOWING": TokenType.FOLLOWING,
 544        "FOR": TokenType.FOR,
 545        "FOREIGN KEY": TokenType.FOREIGN_KEY,
 546        "FORMAT": TokenType.FORMAT,
 547        "FROM": TokenType.FROM,
 548        "GLOB": TokenType.GLOB,
 549        "GROUP BY": TokenType.GROUP_BY,
 550        "GROUPING SETS": TokenType.GROUPING_SETS,
 551        "HAVING": TokenType.HAVING,
 552        "IF": TokenType.IF,
 553        "ILIKE": TokenType.ILIKE,
 554        "IGNORE NULLS": TokenType.IGNORE_NULLS,
 555        "IN": TokenType.IN,
 556        "INDEX": TokenType.INDEX,
 557        "INET": TokenType.INET,
 558        "INNER": TokenType.INNER,
 559        "INSERT": TokenType.INSERT,
 560        "INTERVAL": TokenType.INTERVAL,
 561        "INTERSECT": TokenType.INTERSECT,
 562        "INTO": TokenType.INTO,
 563        "IS": TokenType.IS,
 564        "ISNULL": TokenType.ISNULL,
 565        "JOIN": TokenType.JOIN,
 566        "KEEP": TokenType.KEEP,
 567        "LATERAL": TokenType.LATERAL,
 568        "LAZY": TokenType.LAZY,
 569        "LEADING": TokenType.LEADING,
 570        "LEFT": TokenType.LEFT,
 571        "LIKE": TokenType.LIKE,
 572        "LIMIT": TokenType.LIMIT,
 573        "LOAD DATA": TokenType.LOAD_DATA,
 574        "LOCAL": TokenType.LOCAL,
 575        "MATERIALIZED": TokenType.MATERIALIZED,
 576        "MERGE": TokenType.MERGE,
 577        "NATURAL": TokenType.NATURAL,
 578        "NEXT": TokenType.NEXT,
 579        "NEXT VALUE FOR": TokenType.NEXT_VALUE_FOR,
 580        "NO ACTION": TokenType.NO_ACTION,
 581        "NOT": TokenType.NOT,
 582        "NOTNULL": TokenType.NOTNULL,
 583        "NULL": TokenType.NULL,
 584        "NULLS FIRST": TokenType.NULLS_FIRST,
 585        "NULLS LAST": TokenType.NULLS_LAST,
 586        "OBJECT": TokenType.OBJECT,
 587        "OFFSET": TokenType.OFFSET,
 588        "ON": TokenType.ON,
 589        "ONLY": TokenType.ONLY,
 590        "OPTIONS": TokenType.OPTIONS,
 591        "OR": TokenType.OR,
 592        "ORDER BY": TokenType.ORDER_BY,
 593        "ORDINALITY": TokenType.ORDINALITY,
 594        "OUTER": TokenType.OUTER,
 595        "OUT OF": TokenType.OUT_OF,
 596        "OVER": TokenType.OVER,
 597        "OVERLAPS": TokenType.OVERLAPS,
 598        "OVERWRITE": TokenType.OVERWRITE,
 599        "PARTITION": TokenType.PARTITION,
 600        "PARTITION BY": TokenType.PARTITION_BY,
 601        "PARTITIONED BY": TokenType.PARTITION_BY,
 602        "PARTITIONED_BY": TokenType.PARTITION_BY,
 603        "PERCENT": TokenType.PERCENT,
 604        "PIVOT": TokenType.PIVOT,
 605        "PRAGMA": TokenType.PRAGMA,
 606        "PRECEDING": TokenType.PRECEDING,
 607        "PRIMARY KEY": TokenType.PRIMARY_KEY,
 608        "PROCEDURE": TokenType.PROCEDURE,
 609        "QUALIFY": TokenType.QUALIFY,
 610        "RANGE": TokenType.RANGE,
 611        "RECURSIVE": TokenType.RECURSIVE,
 612        "REGEXP": TokenType.RLIKE,
 613        "REPLACE": TokenType.REPLACE,
 614        "RESPECT NULLS": TokenType.RESPECT_NULLS,
 615        "REFERENCES": TokenType.REFERENCES,
 616        "RIGHT": TokenType.RIGHT,
 617        "RLIKE": TokenType.RLIKE,
 618        "ROLLBACK": TokenType.ROLLBACK,
 619        "ROLLUP": TokenType.ROLLUP,
 620        "ROW": TokenType.ROW,
 621        "ROWS": TokenType.ROWS,
 622        "SCHEMA": TokenType.SCHEMA,
 623        "SEED": TokenType.SEED,
 624        "SELECT": TokenType.SELECT,
 625        "SEMI": TokenType.SEMI,
 626        "SET": TokenType.SET,
 627        "SHOW": TokenType.SHOW,
 628        "SIMILAR TO": TokenType.SIMILAR_TO,
 629        "SOME": TokenType.SOME,
 630        "SORTKEY": TokenType.SORTKEY,
 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        "TRAILING": TokenType.TRAILING,
 639        "UNBOUNDED": TokenType.UNBOUNDED,
 640        "UNION": TokenType.UNION,
 641        "UNLOGGED": TokenType.UNLOGGED,
 642        "UNNEST": TokenType.UNNEST,
 643        "UNPIVOT": TokenType.UNPIVOT,
 644        "UPDATE": TokenType.UPDATE,
 645        "USE": TokenType.USE,
 646        "USING": TokenType.USING,
 647        "UUID": TokenType.UUID,
 648        "VALUES": TokenType.VALUES,
 649        "VIEW": TokenType.VIEW,
 650        "VOLATILE": TokenType.VOLATILE,
 651        "WHEN": TokenType.WHEN,
 652        "WHERE": TokenType.WHERE,
 653        "WINDOW": TokenType.WINDOW,
 654        "WITH": TokenType.WITH,
 655        "WITH TIME ZONE": TokenType.WITH_TIME_ZONE,
 656        "WITH LOCAL TIME ZONE": TokenType.WITH_LOCAL_TIME_ZONE,
 657        "WITHIN GROUP": TokenType.WITHIN_GROUP,
 658        "WITHOUT TIME ZONE": TokenType.WITHOUT_TIME_ZONE,
 659        "APPLY": TokenType.APPLY,
 660        "ARRAY": TokenType.ARRAY,
 661        "BIT": TokenType.BIT,
 662        "BOOL": TokenType.BOOLEAN,
 663        "BOOLEAN": TokenType.BOOLEAN,
 664        "BYTE": TokenType.TINYINT,
 665        "TINYINT": TokenType.TINYINT,
 666        "SHORT": TokenType.SMALLINT,
 667        "SMALLINT": TokenType.SMALLINT,
 668        "INT2": TokenType.SMALLINT,
 669        "INTEGER": TokenType.INT,
 670        "INT": TokenType.INT,
 671        "INT4": TokenType.INT,
 672        "LONG": TokenType.BIGINT,
 673        "BIGINT": TokenType.BIGINT,
 674        "INT8": TokenType.BIGINT,
 675        "DEC": TokenType.DECIMAL,
 676        "DECIMAL": TokenType.DECIMAL,
 677        "BIGDECIMAL": TokenType.BIGDECIMAL,
 678        "BIGNUMERIC": TokenType.BIGDECIMAL,
 679        "MAP": TokenType.MAP,
 680        "NULLABLE": TokenType.NULLABLE,
 681        "NUMBER": TokenType.DECIMAL,
 682        "NUMERIC": TokenType.DECIMAL,
 683        "FIXED": TokenType.DECIMAL,
 684        "REAL": TokenType.FLOAT,
 685        "FLOAT": TokenType.FLOAT,
 686        "FLOAT4": TokenType.FLOAT,
 687        "FLOAT8": TokenType.DOUBLE,
 688        "DOUBLE": TokenType.DOUBLE,
 689        "DOUBLE PRECISION": TokenType.DOUBLE,
 690        "JSON": TokenType.JSON,
 691        "CHAR": TokenType.CHAR,
 692        "CHARACTER": TokenType.CHAR,
 693        "NCHAR": TokenType.NCHAR,
 694        "VARCHAR": TokenType.VARCHAR,
 695        "VARCHAR2": TokenType.VARCHAR,
 696        "NVARCHAR": TokenType.NVARCHAR,
 697        "NVARCHAR2": TokenType.NVARCHAR,
 698        "STR": TokenType.TEXT,
 699        "STRING": TokenType.TEXT,
 700        "TEXT": TokenType.TEXT,
 701        "CLOB": TokenType.TEXT,
 702        "LONGVARCHAR": TokenType.TEXT,
 703        "BINARY": TokenType.BINARY,
 704        "BLOB": TokenType.VARBINARY,
 705        "BYTEA": TokenType.VARBINARY,
 706        "VARBINARY": TokenType.VARBINARY,
 707        "TIME": TokenType.TIME,
 708        "TIMESTAMP": TokenType.TIMESTAMP,
 709        "TIMESTAMPTZ": TokenType.TIMESTAMPTZ,
 710        "TIMESTAMPLTZ": TokenType.TIMESTAMPLTZ,
 711        "DATE": TokenType.DATE,
 712        "DATETIME": TokenType.DATETIME,
 713        "UNIQUE": TokenType.UNIQUE,
 714        "STRUCT": TokenType.STRUCT,
 715        "VARIANT": TokenType.VARIANT,
 716        "ALTER": TokenType.ALTER,
 717        "ALTER AGGREGATE": TokenType.COMMAND,
 718        "ALTER DEFAULT": TokenType.COMMAND,
 719        "ALTER DOMAIN": TokenType.COMMAND,
 720        "ALTER ROLE": TokenType.COMMAND,
 721        "ALTER RULE": TokenType.COMMAND,
 722        "ALTER SEQUENCE": TokenType.COMMAND,
 723        "ALTER TYPE": TokenType.COMMAND,
 724        "ALTER USER": TokenType.COMMAND,
 725        "ALTER VIEW": TokenType.COMMAND,
 726        "ANALYZE": TokenType.COMMAND,
 727        "CALL": TokenType.COMMAND,
 728        "COMMENT": TokenType.COMMENT,
 729        "COPY": TokenType.COMMAND,
 730        "EXPLAIN": TokenType.COMMAND,
 731        "GRANT": TokenType.COMMAND,
 732        "OPTIMIZE": TokenType.COMMAND,
 733        "PREPARE": TokenType.COMMAND,
 734        "TRUNCATE": TokenType.COMMAND,
 735        "VACUUM": TokenType.COMMAND,
 736    }
 737
 738    WHITE_SPACE: t.Dict[t.Optional[str], TokenType] = {
 739        " ": TokenType.SPACE,
 740        "\t": TokenType.SPACE,
 741        "\n": TokenType.BREAK,
 742        "\r": TokenType.BREAK,
 743        "\r\n": TokenType.BREAK,
 744    }
 745
 746    COMMANDS = {
 747        TokenType.COMMAND,
 748        TokenType.EXECUTE,
 749        TokenType.FETCH,
 750        TokenType.SHOW,
 751    }
 752
 753    COMMAND_PREFIX_TOKENS = {TokenType.SEMICOLON, TokenType.BEGIN}
 754
 755    # handle numeric literals like in hive (3L = BIGINT)
 756    NUMERIC_LITERALS: t.Dict[str, str] = {}
 757    ENCODE: t.Optional[str] = None
 758
 759    COMMENTS = ["--", ("/*", "*/"), ("{#", "#}")]
 760    KEYWORD_TRIE: t.Dict = {}  # autofilled
 761
 762    IDENTIFIER_CAN_START_WITH_DIGIT = False
 763
 764    __slots__ = (
 765        "sql",
 766        "size",
 767        "tokens",
 768        "_start",
 769        "_current",
 770        "_line",
 771        "_col",
 772        "_comments",
 773        "_char",
 774        "_end",
 775        "_peek",
 776        "_prev_token_line",
 777        "_prev_token_comments",
 778        "_prev_token_type",
 779    )
 780
 781    def __init__(self) -> None:
 782        self.reset()
 783
 784    def reset(self) -> None:
 785        self.sql = ""
 786        self.size = 0
 787        self.tokens: t.List[Token] = []
 788        self._start = 0
 789        self._current = 0
 790        self._line = 1
 791        self._col = 1
 792        self._comments: t.List[str] = []
 793
 794        self._char = ""
 795        self._end = False
 796        self._peek = ""
 797        self._prev_token_line = -1
 798        self._prev_token_comments: t.List[str] = []
 799        self._prev_token_type: t.Optional[TokenType] = None
 800
 801    def tokenize(self, sql: str) -> t.List[Token]:
 802        """Returns a list of tokens corresponding to the SQL string `sql`."""
 803        self.reset()
 804        self.sql = sql
 805        self.size = len(sql)
 806        try:
 807            self._scan()
 808        except Exception as e:
 809            start = self._current - 50
 810            end = self._current + 50
 811            start = start if start > 0 else 0
 812            end = end if end < self.size else self.size - 1
 813            context = self.sql[start:end]
 814            raise ValueError(f"Error tokenizing '{context}'") from e
 815
 816        return self.tokens
 817
 818    def _scan(self, until: t.Optional[t.Callable] = None) -> None:
 819        while self.size and not self._end:
 820            self._start = self._current
 821            self._advance()
 822
 823            if self._char is None:
 824                break
 825
 826            if self._char not in self.WHITE_SPACE:
 827                if self._char.isdigit():
 828                    self._scan_number()
 829                elif self._char in self._IDENTIFIERS:
 830                    self._scan_identifier(self._IDENTIFIERS[self._char])
 831                else:
 832                    self._scan_keywords()
 833
 834            if until and until():
 835                break
 836
 837        if self.tokens:
 838            self.tokens[-1].comments.extend(self._comments)
 839
 840    def _chars(self, size: int) -> str:
 841        if size == 1:
 842            return self._char
 843        start = self._current - 1
 844        end = start + size
 845        if end <= self.size:
 846            return self.sql[start:end]
 847        return ""
 848
 849    def _advance(self, i: int = 1) -> None:
 850        if self.WHITE_SPACE.get(self._char) is TokenType.BREAK:
 851            self._col = 1
 852            self._line += 1
 853        else:
 854            self._col += i
 855
 856        self._current += i
 857        self._end = self._current >= self.size
 858        self._char = self.sql[self._current - 1]
 859        self._peek = "" if self._end else self.sql[self._current]
 860
 861    @property
 862    def _text(self) -> str:
 863        return self.sql[self._start : self._current]
 864
 865    def _add(self, token_type: TokenType, text: t.Optional[str] = None) -> None:
 866        self._prev_token_line = self._line
 867        self._prev_token_comments = self._comments
 868        self._prev_token_type = token_type
 869        self.tokens.append(
 870            Token(
 871                token_type,
 872                self._text if text is None else text,
 873                self._line,
 874                self._col,
 875                self._current,
 876                self._comments,
 877            )
 878        )
 879        self._comments = []
 880
 881        # If we have either a semicolon or a begin token before the command's token, we'll parse
 882        # whatever follows the command's token as a string
 883        if (
 884            token_type in self.COMMANDS
 885            and self._peek != ";"
 886            and (len(self.tokens) == 1 or self.tokens[-2].token_type in self.COMMAND_PREFIX_TOKENS)
 887        ):
 888            start = self._current
 889            tokens = len(self.tokens)
 890            self._scan(lambda: self._peek == ";")
 891            self.tokens = self.tokens[:tokens]
 892            text = self.sql[start : self._current].strip()
 893            if text:
 894                self._add(TokenType.STRING, text)
 895
 896    def _scan_keywords(self) -> None:
 897        size = 0
 898        word = None
 899        chars = self._text
 900        char = chars
 901        prev_space = False
 902        skip = False
 903        trie = self.KEYWORD_TRIE
 904        single_token = char in self.SINGLE_TOKENS
 905
 906        while chars:
 907            if skip:
 908                result = 1
 909            else:
 910                result, trie = in_trie(trie, char.upper())
 911
 912            if result == 0:
 913                break
 914            if result == 2:
 915                word = chars
 916            size += 1
 917            end = self._current - 1 + size
 918
 919            if end < self.size:
 920                char = self.sql[end]
 921                single_token = single_token or char in self.SINGLE_TOKENS
 922                is_space = char in self.WHITE_SPACE
 923
 924                if not is_space or not prev_space:
 925                    if is_space:
 926                        char = " "
 927                    chars += char
 928                    prev_space = is_space
 929                    skip = False
 930                else:
 931                    skip = True
 932            else:
 933                chars = " "
 934
 935        word = None if not single_token and chars[-1] not in self.WHITE_SPACE else word
 936
 937        if not word:
 938            if self._char in self.SINGLE_TOKENS:
 939                self._add(self.SINGLE_TOKENS[self._char], text=self._char)
 940                return
 941            self._scan_var()
 942            return
 943
 944        if self._scan_string(word):
 945            return
 946        if self._scan_formatted_string(word):
 947            return
 948        if self._scan_comment(word):
 949            return
 950
 951        self._advance(size - 1)
 952        word = word.upper()
 953        self._add(self.KEYWORDS[word], text=word)
 954
 955    def _scan_comment(self, comment_start: str) -> bool:
 956        if comment_start not in self._COMMENTS:
 957            return False
 958
 959        comment_start_line = self._line
 960        comment_start_size = len(comment_start)
 961        comment_end = self._COMMENTS[comment_start]
 962
 963        if comment_end:
 964            # Skip the comment's start delimiter
 965            self._advance(comment_start_size)
 966
 967            comment_end_size = len(comment_end)
 968            while not self._end and self._chars(comment_end_size) != comment_end:
 969                self._advance()
 970
 971            self._comments.append(self._text[comment_start_size : -comment_end_size + 1])
 972            self._advance(comment_end_size - 1)
 973        else:
 974            while not self._end and not self.WHITE_SPACE.get(self._peek) is TokenType.BREAK:
 975                self._advance()
 976            self._comments.append(self._text[comment_start_size:])
 977
 978        # Leading comment is attached to the succeeding token, whilst trailing comment to the preceding.
 979        # Multiple consecutive comments are preserved by appending them to the current comments list.
 980        if comment_start_line == self._prev_token_line:
 981            self.tokens[-1].comments.extend(self._comments)
 982            self._comments = []
 983            self._prev_token_line = self._line
 984
 985        return True
 986
 987    def _scan_number(self) -> None:
 988        if self._char == "0":
 989            peek = self._peek.upper()
 990            if peek == "B":
 991                return self._scan_bits()
 992            elif peek == "X":
 993                return self._scan_hex()
 994
 995        decimal = False
 996        scientific = 0
 997
 998        while True:
 999            if self._peek.isdigit():
1000                self._advance()
1001            elif self._peek == "." and not decimal:
1002                decimal = True
1003                self._advance()
1004            elif self._peek in ("-", "+") and scientific == 1:
1005                scientific += 1
1006                self._advance()
1007            elif self._peek.upper() == "E" and not scientific:
1008                scientific += 1
1009                self._advance()
1010            elif self._peek.isidentifier():
1011                number_text = self._text
1012                literal = ""
1013
1014                while self._peek.strip() and self._peek not in self.SINGLE_TOKENS:
1015                    literal += self._peek.upper()
1016                    self._advance()
1017
1018                token_type = self.KEYWORDS.get(self.NUMERIC_LITERALS.get(literal))
1019
1020                if token_type:
1021                    self._add(TokenType.NUMBER, number_text)
1022                    self._add(TokenType.DCOLON, "::")
1023                    return self._add(token_type, literal)
1024                elif self.IDENTIFIER_CAN_START_WITH_DIGIT:
1025                    return self._add(TokenType.VAR)
1026
1027                self._add(TokenType.NUMBER, number_text)
1028                return self._advance(-len(literal))
1029            else:
1030                return self._add(TokenType.NUMBER)
1031
1032    def _scan_bits(self) -> None:
1033        self._advance()
1034        value = self._extract_value()
1035        try:
1036            self._add(TokenType.BIT_STRING, f"{int(value, 2)}")
1037        except ValueError:
1038            self._add(TokenType.IDENTIFIER)
1039
1040    def _scan_hex(self) -> None:
1041        self._advance()
1042        value = self._extract_value()
1043        try:
1044            self._add(TokenType.HEX_STRING, f"{int(value, 16)}")
1045        except ValueError:
1046            self._add(TokenType.IDENTIFIER)
1047
1048    def _extract_value(self) -> str:
1049        while True:
1050            char = self._peek.strip()
1051            if char and char not in self.SINGLE_TOKENS:
1052                self._advance()
1053            else:
1054                break
1055
1056        return self._text
1057
1058    def _scan_string(self, quote: str) -> bool:
1059        quote_end = self._QUOTES.get(quote)
1060        if quote_end is None:
1061            return False
1062
1063        self._advance(len(quote))
1064        text = self._extract_string(quote_end)
1065        text = text.encode(self.ENCODE).decode(self.ENCODE) if self.ENCODE else text
1066        self._add(TokenType.NATIONAL if quote[0].upper() == "N" else TokenType.STRING, text)
1067        return True
1068
1069    # X'1234, b'0110', E'\\\\\' etc.
1070    def _scan_formatted_string(self, string_start: str) -> bool:
1071        if string_start in self._HEX_STRINGS:
1072            delimiters = self._HEX_STRINGS
1073            token_type = TokenType.HEX_STRING
1074            base = 16
1075        elif string_start in self._BIT_STRINGS:
1076            delimiters = self._BIT_STRINGS
1077            token_type = TokenType.BIT_STRING
1078            base = 2
1079        elif string_start in self._BYTE_STRINGS:
1080            delimiters = self._BYTE_STRINGS
1081            token_type = TokenType.BYTE_STRING
1082            base = None
1083        else:
1084            return False
1085
1086        self._advance(len(string_start))
1087        string_end = delimiters[string_start]
1088        text = self._extract_string(string_end)
1089
1090        if base is None:
1091            self._add(token_type, text)
1092        else:
1093            try:
1094                self._add(token_type, f"{int(text, base)}")
1095            except:
1096                raise RuntimeError(
1097                    f"Numeric string contains invalid characters from {self._line}:{self._start}"
1098                )
1099
1100        return True
1101
1102    def _scan_identifier(self, identifier_end: str) -> None:
1103        text = ""
1104        identifier_end_is_escape = identifier_end in self._IDENTIFIER_ESCAPES
1105
1106        while True:
1107            if self._end:
1108                raise RuntimeError(f"Missing {identifier_end} from {self._line}:{self._start}")
1109
1110            self._advance()
1111            if self._char == identifier_end:
1112                if identifier_end_is_escape and self._peek == identifier_end:
1113                    text += identifier_end
1114                    self._advance()
1115                    continue
1116
1117                break
1118
1119            text += self._char
1120
1121        self._add(TokenType.IDENTIFIER, text)
1122
1123    def _scan_var(self) -> None:
1124        while True:
1125            char = self._peek.strip()
1126            if char and (char in self.VAR_SINGLE_TOKENS or char not in self.SINGLE_TOKENS):
1127                self._advance()
1128            else:
1129                break
1130        self._add(
1131            TokenType.VAR
1132            if self._prev_token_type == TokenType.PARAMETER
1133            else self.KEYWORDS.get(self._text.upper(), TokenType.VAR)
1134        )
1135
1136    def _extract_string(self, delimiter: str) -> str:
1137        text = ""
1138        delim_size = len(delimiter)
1139
1140        while True:
1141            if self._char in self._STRING_ESCAPES and (
1142                self._peek == delimiter or self._peek in self._STRING_ESCAPES
1143            ):
1144                if self._peek == delimiter:
1145                    text += self._peek
1146                else:
1147                    text += self._char + self._peek
1148
1149                if self._current + 1 < self.size:
1150                    self._advance(2)
1151                else:
1152                    raise RuntimeError(f"Missing {delimiter} from {self._line}:{self._current}")
1153            else:
1154                if self._chars(delim_size) == delimiter:
1155                    if delim_size > 1:
1156                        self._advance(delim_size - 1)
1157                    break
1158
1159                if self._end:
1160                    raise RuntimeError(f"Missing {delimiter} from {self._line}:{self._start}")
1161                text += self._char
1162                self._advance()
1163
1164        return text