sqlglot.tokens

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