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