sqlglot.tokens

   1from __future__ import annotations
   2
   3import typing as t
   4from enum import auto
   5
   6from sqlglot.errors import TokenError
   7from sqlglot.helper import AutoName
   8from sqlglot.trie import TrieResult, in_trie, new_trie
   9
  10
  11class TokenType(AutoName):
  12    L_PAREN = auto()
  13    R_PAREN = auto()
  14    L_BRACKET = auto()
  15    R_BRACKET = auto()
  16    L_BRACE = auto()
  17    R_BRACE = auto()
  18    COMMA = auto()
  19    DOT = auto()
  20    DASH = auto()
  21    PLUS = auto()
  22    COLON = auto()
  23    DCOLON = auto()
  24    DQMARK = auto()
  25    SEMICOLON = auto()
  26    STAR = auto()
  27    BACKSLASH = auto()
  28    SLASH = auto()
  29    LT = auto()
  30    LTE = auto()
  31    GT = auto()
  32    GTE = auto()
  33    NOT = auto()
  34    EQ = auto()
  35    NEQ = auto()
  36    NULLSAFE_EQ = auto()
  37    AND = auto()
  38    OR = auto()
  39    AMP = auto()
  40    DPIPE = auto()
  41    PIPE = auto()
  42    CARET = auto()
  43    TILDA = auto()
  44    ARROW = auto()
  45    DARROW = auto()
  46    FARROW = auto()
  47    HASH = auto()
  48    HASH_ARROW = auto()
  49    DHASH_ARROW = auto()
  50    LR_ARROW = auto()
  51    DAT = auto()
  52    LT_AT = auto()
  53    AT_GT = auto()
  54    DOLLAR = auto()
  55    PARAMETER = auto()
  56    SESSION_PARAMETER = auto()
  57    DAMP = auto()
  58    XOR = auto()
  59
  60    BLOCK_START = auto()
  61    BLOCK_END = auto()
  62
  63    SPACE = auto()
  64    BREAK = auto()
  65
  66    STRING = auto()
  67    NUMBER = auto()
  68    IDENTIFIER = auto()
  69    DATABASE = auto()
  70    COLUMN = auto()
  71    COLUMN_DEF = auto()
  72    SCHEMA = auto()
  73    TABLE = auto()
  74    VAR = auto()
  75    BIT_STRING = auto()
  76    HEX_STRING = auto()
  77    BYTE_STRING = auto()
  78    NATIONAL_STRING = auto()
  79    RAW_STRING = auto()
  80
  81    # types
  82    BIT = auto()
  83    BOOLEAN = auto()
  84    TINYINT = auto()
  85    UTINYINT = auto()
  86    SMALLINT = auto()
  87    USMALLINT = auto()
  88    MEDIUMINT = auto()
  89    UMEDIUMINT = auto()
  90    INT = auto()
  91    UINT = auto()
  92    BIGINT = auto()
  93    UBIGINT = auto()
  94    INT128 = auto()
  95    UINT128 = auto()
  96    INT256 = auto()
  97    UINT256 = auto()
  98    FLOAT = auto()
  99    DOUBLE = auto()
 100    DECIMAL = auto()
 101    BIGDECIMAL = auto()
 102    CHAR = auto()
 103    NCHAR = auto()
 104    VARCHAR = auto()
 105    NVARCHAR = auto()
 106    TEXT = auto()
 107    MEDIUMTEXT = auto()
 108    LONGTEXT = auto()
 109    MEDIUMBLOB = auto()
 110    LONGBLOB = auto()
 111    TINYBLOB = auto()
 112    TINYTEXT = auto()
 113    BINARY = auto()
 114    VARBINARY = auto()
 115    JSON = auto()
 116    JSONB = auto()
 117    TIME = auto()
 118    TIMETZ = auto()
 119    TIMESTAMP = auto()
 120    TIMESTAMPTZ = auto()
 121    TIMESTAMPLTZ = auto()
 122    DATETIME = auto()
 123    DATETIME64 = auto()
 124    DATE = auto()
 125    INT4RANGE = auto()
 126    INT4MULTIRANGE = auto()
 127    INT8RANGE = auto()
 128    INT8MULTIRANGE = auto()
 129    NUMRANGE = auto()
 130    NUMMULTIRANGE = auto()
 131    TSRANGE = auto()
 132    TSMULTIRANGE = auto()
 133    TSTZRANGE = auto()
 134    TSTZMULTIRANGE = auto()
 135    DATERANGE = auto()
 136    DATEMULTIRANGE = auto()
 137    UUID = auto()
 138    GEOGRAPHY = auto()
 139    NULLABLE = auto()
 140    GEOMETRY = auto()
 141    HLLSKETCH = auto()
 142    HSTORE = auto()
 143    SUPER = auto()
 144    SERIAL = auto()
 145    SMALLSERIAL = auto()
 146    BIGSERIAL = auto()
 147    XML = auto()
 148    YEAR = auto()
 149    UNIQUEIDENTIFIER = auto()
 150    USERDEFINED = auto()
 151    MONEY = auto()
 152    SMALLMONEY = auto()
 153    ROWVERSION = auto()
 154    IMAGE = auto()
 155    VARIANT = auto()
 156    OBJECT = auto()
 157    INET = auto()
 158    IPADDRESS = auto()
 159    IPPREFIX = auto()
 160    ENUM = auto()
 161    ENUM8 = auto()
 162    ENUM16 = auto()
 163    FIXEDSTRING = auto()
 164    LOWCARDINALITY = auto()
 165    NESTED = auto()
 166    UNKNOWN = auto()
 167
 168    # keywords
 169    ALIAS = auto()
 170    ALTER = auto()
 171    ALWAYS = auto()
 172    ALL = auto()
 173    ANTI = auto()
 174    ANY = auto()
 175    APPLY = auto()
 176    ARRAY = auto()
 177    ASC = auto()
 178    ASOF = auto()
 179    AUTO_INCREMENT = auto()
 180    BEGIN = auto()
 181    BETWEEN = auto()
 182    CACHE = auto()
 183    CASE = auto()
 184    CHARACTER_SET = auto()
 185    CLUSTER_BY = auto()
 186    COLLATE = auto()
 187    COMMAND = auto()
 188    COMMENT = auto()
 189    COMMIT = auto()
 190    CONNECT_BY = auto()
 191    CONSTRAINT = auto()
 192    CREATE = auto()
 193    CROSS = auto()
 194    CUBE = auto()
 195    CURRENT_DATE = auto()
 196    CURRENT_DATETIME = auto()
 197    CURRENT_TIME = auto()
 198    CURRENT_TIMESTAMP = auto()
 199    CURRENT_USER = auto()
 200    DEFAULT = auto()
 201    DELETE = auto()
 202    DESC = auto()
 203    DESCRIBE = auto()
 204    DICTIONARY = auto()
 205    DISTINCT = auto()
 206    DISTRIBUTE_BY = auto()
 207    DIV = auto()
 208    DROP = auto()
 209    ELSE = auto()
 210    END = auto()
 211    ESCAPE = auto()
 212    EXCEPT = auto()
 213    EXECUTE = auto()
 214    EXISTS = auto()
 215    FALSE = auto()
 216    FETCH = auto()
 217    FILTER = auto()
 218    FINAL = auto()
 219    FIRST = auto()
 220    FOR = auto()
 221    FORCE = auto()
 222    FOREIGN_KEY = auto()
 223    FORMAT = auto()
 224    FROM = auto()
 225    FULL = auto()
 226    FUNCTION = auto()
 227    GLOB = auto()
 228    GLOBAL = auto()
 229    GROUP_BY = auto()
 230    GROUPING_SETS = auto()
 231    HAVING = auto()
 232    HINT = auto()
 233    IGNORE = auto()
 234    ILIKE = auto()
 235    ILIKE_ANY = auto()
 236    IN = auto()
 237    INDEX = auto()
 238    INNER = auto()
 239    INSERT = auto()
 240    INTERSECT = auto()
 241    INTERVAL = auto()
 242    INTO = auto()
 243    INTRODUCER = auto()
 244    IRLIKE = auto()
 245    IS = auto()
 246    ISNULL = auto()
 247    JOIN = auto()
 248    JOIN_MARKER = auto()
 249    KEEP = auto()
 250    LANGUAGE = auto()
 251    LATERAL = auto()
 252    LEFT = auto()
 253    LIKE = auto()
 254    LIKE_ANY = auto()
 255    LIMIT = auto()
 256    LOAD = auto()
 257    LOCK = auto()
 258    MAP = auto()
 259    MATCH_RECOGNIZE = auto()
 260    MEMBER_OF = auto()
 261    MERGE = auto()
 262    MOD = auto()
 263    NATURAL = auto()
 264    NEXT = auto()
 265    NOTNULL = auto()
 266    NULL = auto()
 267    OBJECT_IDENTIFIER = auto()
 268    OFFSET = auto()
 269    ON = auto()
 270    ORDER_BY = auto()
 271    ORDERED = auto()
 272    ORDINALITY = auto()
 273    OUTER = auto()
 274    OVER = auto()
 275    OVERLAPS = auto()
 276    OVERWRITE = auto()
 277    PARTITION = auto()
 278    PARTITION_BY = auto()
 279    PERCENT = auto()
 280    PIVOT = auto()
 281    PLACEHOLDER = auto()
 282    PRAGMA = auto()
 283    PRIMARY_KEY = auto()
 284    PROCEDURE = auto()
 285    PROPERTIES = auto()
 286    PSEUDO_TYPE = auto()
 287    QUALIFY = auto()
 288    QUOTE = auto()
 289    RANGE = auto()
 290    RECURSIVE = auto()
 291    REPLACE = auto()
 292    RETURNING = auto()
 293    REFERENCES = auto()
 294    RIGHT = auto()
 295    RLIKE = auto()
 296    ROLLBACK = auto()
 297    ROLLUP = auto()
 298    ROW = auto()
 299    ROWS = auto()
 300    SELECT = auto()
 301    SEMI = auto()
 302    SEPARATOR = auto()
 303    SERDE_PROPERTIES = auto()
 304    SET = auto()
 305    SETTINGS = auto()
 306    SHOW = auto()
 307    SIMILAR_TO = auto()
 308    SOME = auto()
 309    SORT_BY = auto()
 310    START_WITH = auto()
 311    STRUCT = auto()
 312    TABLE_SAMPLE = auto()
 313    TEMPORARY = auto()
 314    TOP = auto()
 315    THEN = auto()
 316    TRUE = auto()
 317    UNCACHE = auto()
 318    UNION = auto()
 319    UNNEST = auto()
 320    UNPIVOT = auto()
 321    UPDATE = auto()
 322    USE = auto()
 323    USING = auto()
 324    VALUES = auto()
 325    VIEW = auto()
 326    VOLATILE = auto()
 327    WHEN = auto()
 328    WHERE = auto()
 329    WINDOW = auto()
 330    WITH = auto()
 331    UNIQUE = auto()
 332    VERSION_SNAPSHOT = auto()
 333    TIMESTAMP_SNAPSHOT = auto()
 334
 335
 336class Token:
 337    __slots__ = ("token_type", "text", "line", "col", "start", "end", "comments")
 338
 339    @classmethod
 340    def number(cls, number: int) -> Token:
 341        """Returns a NUMBER token with `number` as its text."""
 342        return cls(TokenType.NUMBER, str(number))
 343
 344    @classmethod
 345    def string(cls, string: str) -> Token:
 346        """Returns a STRING token with `string` as its text."""
 347        return cls(TokenType.STRING, string)
 348
 349    @classmethod
 350    def identifier(cls, identifier: str) -> Token:
 351        """Returns an IDENTIFIER token with `identifier` as its text."""
 352        return cls(TokenType.IDENTIFIER, identifier)
 353
 354    @classmethod
 355    def var(cls, var: str) -> Token:
 356        """Returns an VAR token with `var` as its text."""
 357        return cls(TokenType.VAR, var)
 358
 359    def __init__(
 360        self,
 361        token_type: TokenType,
 362        text: str,
 363        line: int = 1,
 364        col: int = 1,
 365        start: int = 0,
 366        end: int = 0,
 367        comments: t.List[str] = [],
 368    ) -> None:
 369        """Token initializer.
 370
 371        Args:
 372            token_type: The TokenType Enum.
 373            text: The text of the token.
 374            line: The line that the token ends on.
 375            col: The column that the token ends on.
 376            start: The start index of the token.
 377            end: The ending index of the token.
 378            comments: The comments to attach to the token.
 379        """
 380        self.token_type = token_type
 381        self.text = text
 382        self.line = line
 383        self.col = col
 384        self.start = start
 385        self.end = end
 386        self.comments = comments
 387
 388    def __repr__(self) -> str:
 389        attributes = ", ".join(f"{k}: {getattr(self, k)}" for k in self.__slots__)
 390        return f"<Token {attributes}>"
 391
 392
 393class _Tokenizer(type):
 394    def __new__(cls, clsname, bases, attrs):
 395        klass = super().__new__(cls, clsname, bases, attrs)
 396
 397        def _convert_quotes(arr: t.List[str | t.Tuple[str, str]]) -> t.Dict[str, str]:
 398            return dict(
 399                (item, item) if isinstance(item, str) else (item[0], item[1]) for item in arr
 400            )
 401
 402        def _quotes_to_format(
 403            token_type: TokenType, arr: t.List[str | t.Tuple[str, str]]
 404        ) -> t.Dict[str, t.Tuple[str, TokenType]]:
 405            return {k: (v, token_type) for k, v in _convert_quotes(arr).items()}
 406
 407        klass._QUOTES = _convert_quotes(klass.QUOTES)
 408        klass._IDENTIFIERS = _convert_quotes(klass.IDENTIFIERS)
 409
 410        klass._FORMAT_STRINGS = {
 411            **{
 412                p + s: (e, TokenType.NATIONAL_STRING)
 413                for s, e in klass._QUOTES.items()
 414                for p in ("n", "N")
 415            },
 416            **_quotes_to_format(TokenType.BIT_STRING, klass.BIT_STRINGS),
 417            **_quotes_to_format(TokenType.BYTE_STRING, klass.BYTE_STRINGS),
 418            **_quotes_to_format(TokenType.HEX_STRING, klass.HEX_STRINGS),
 419            **_quotes_to_format(TokenType.RAW_STRING, klass.RAW_STRINGS),
 420        }
 421
 422        klass._STRING_ESCAPES = set(klass.STRING_ESCAPES)
 423        klass._IDENTIFIER_ESCAPES = set(klass.IDENTIFIER_ESCAPES)
 424        klass._COMMENTS = {
 425            **dict(
 426                (comment, None) if isinstance(comment, str) else (comment[0], comment[1])
 427                for comment in klass.COMMENTS
 428            ),
 429            "{#": "#}",  # Ensure Jinja comments are tokenized correctly in all dialects
 430        }
 431
 432        klass._KEYWORD_TRIE = new_trie(
 433            key.upper()
 434            for key in (
 435                *klass.KEYWORDS,
 436                *klass._COMMENTS,
 437                *klass._QUOTES,
 438                *klass._FORMAT_STRINGS,
 439            )
 440            if " " in key or any(single in key for single in klass.SINGLE_TOKENS)
 441        )
 442
 443        return klass
 444
 445
 446class Tokenizer(metaclass=_Tokenizer):
 447    SINGLE_TOKENS = {
 448        "(": TokenType.L_PAREN,
 449        ")": TokenType.R_PAREN,
 450        "[": TokenType.L_BRACKET,
 451        "]": TokenType.R_BRACKET,
 452        "{": TokenType.L_BRACE,
 453        "}": TokenType.R_BRACE,
 454        "&": TokenType.AMP,
 455        "^": TokenType.CARET,
 456        ":": TokenType.COLON,
 457        ",": TokenType.COMMA,
 458        ".": TokenType.DOT,
 459        "-": TokenType.DASH,
 460        "=": TokenType.EQ,
 461        ">": TokenType.GT,
 462        "<": TokenType.LT,
 463        "%": TokenType.MOD,
 464        "!": TokenType.NOT,
 465        "|": TokenType.PIPE,
 466        "+": TokenType.PLUS,
 467        ";": TokenType.SEMICOLON,
 468        "/": TokenType.SLASH,
 469        "\\": TokenType.BACKSLASH,
 470        "*": TokenType.STAR,
 471        "~": TokenType.TILDA,
 472        "?": TokenType.PLACEHOLDER,
 473        "@": TokenType.PARAMETER,
 474        # used for breaking a var like x'y' but nothing else
 475        # the token type doesn't matter
 476        "'": TokenType.QUOTE,
 477        "`": TokenType.IDENTIFIER,
 478        '"': TokenType.IDENTIFIER,
 479        "#": TokenType.HASH,
 480    }
 481
 482    BIT_STRINGS: t.List[str | t.Tuple[str, str]] = []
 483    BYTE_STRINGS: t.List[str | t.Tuple[str, str]] = []
 484    HEX_STRINGS: t.List[str | t.Tuple[str, str]] = []
 485    RAW_STRINGS: t.List[str | t.Tuple[str, str]] = []
 486    IDENTIFIERS: t.List[str | t.Tuple[str, str]] = ['"']
 487    IDENTIFIER_ESCAPES = ['"']
 488    QUOTES: t.List[t.Tuple[str, str] | str] = ["'"]
 489    STRING_ESCAPES = ["'"]
 490    VAR_SINGLE_TOKENS: t.Set[str] = set()
 491
 492    # Autofilled
 493    IDENTIFIERS_CAN_START_WITH_DIGIT: bool = False
 494
 495    _COMMENTS: t.Dict[str, str] = {}
 496    _FORMAT_STRINGS: t.Dict[str, t.Tuple[str, TokenType]] = {}
 497    _IDENTIFIERS: t.Dict[str, str] = {}
 498    _IDENTIFIER_ESCAPES: t.Set[str] = set()
 499    _QUOTES: t.Dict[str, str] = {}
 500    _STRING_ESCAPES: t.Set[str] = set()
 501    _KEYWORD_TRIE: t.Dict = {}
 502
 503    KEYWORDS: t.Dict[str, TokenType] = {
 504        **{f"{{%{postfix}": TokenType.BLOCK_START for postfix in ("", "+", "-")},
 505        **{f"{prefix}%}}": TokenType.BLOCK_END for prefix in ("", "+", "-")},
 506        **{f"{{{{{postfix}": TokenType.BLOCK_START for postfix in ("+", "-")},
 507        **{f"{prefix}}}}}": TokenType.BLOCK_END for prefix in ("+", "-")},
 508        "/*+": TokenType.HINT,
 509        "==": TokenType.EQ,
 510        "::": TokenType.DCOLON,
 511        "||": TokenType.DPIPE,
 512        ">=": TokenType.GTE,
 513        "<=": TokenType.LTE,
 514        "<>": TokenType.NEQ,
 515        "!=": TokenType.NEQ,
 516        "<=>": TokenType.NULLSAFE_EQ,
 517        "->": TokenType.ARROW,
 518        "->>": TokenType.DARROW,
 519        "=>": TokenType.FARROW,
 520        "#>": TokenType.HASH_ARROW,
 521        "#>>": TokenType.DHASH_ARROW,
 522        "<->": TokenType.LR_ARROW,
 523        "&&": TokenType.DAMP,
 524        "??": TokenType.DQMARK,
 525        "ALL": TokenType.ALL,
 526        "ALWAYS": TokenType.ALWAYS,
 527        "AND": TokenType.AND,
 528        "ANTI": TokenType.ANTI,
 529        "ANY": TokenType.ANY,
 530        "ASC": TokenType.ASC,
 531        "AS": TokenType.ALIAS,
 532        "ASOF": TokenType.ASOF,
 533        "AUTOINCREMENT": TokenType.AUTO_INCREMENT,
 534        "AUTO_INCREMENT": TokenType.AUTO_INCREMENT,
 535        "BEGIN": TokenType.BEGIN,
 536        "BETWEEN": TokenType.BETWEEN,
 537        "CACHE": TokenType.CACHE,
 538        "UNCACHE": TokenType.UNCACHE,
 539        "CASE": TokenType.CASE,
 540        "CHARACTER SET": TokenType.CHARACTER_SET,
 541        "CLUSTER BY": TokenType.CLUSTER_BY,
 542        "COLLATE": TokenType.COLLATE,
 543        "COLUMN": TokenType.COLUMN,
 544        "COMMIT": TokenType.COMMIT,
 545        "CONNECT BY": TokenType.CONNECT_BY,
 546        "CONSTRAINT": TokenType.CONSTRAINT,
 547        "CREATE": TokenType.CREATE,
 548        "CROSS": TokenType.CROSS,
 549        "CUBE": TokenType.CUBE,
 550        "CURRENT_DATE": TokenType.CURRENT_DATE,
 551        "CURRENT_TIME": TokenType.CURRENT_TIME,
 552        "CURRENT_TIMESTAMP": TokenType.CURRENT_TIMESTAMP,
 553        "CURRENT_USER": TokenType.CURRENT_USER,
 554        "DATABASE": TokenType.DATABASE,
 555        "DEFAULT": TokenType.DEFAULT,
 556        "DELETE": TokenType.DELETE,
 557        "DESC": TokenType.DESC,
 558        "DESCRIBE": TokenType.DESCRIBE,
 559        "DISTINCT": TokenType.DISTINCT,
 560        "DISTRIBUTE BY": TokenType.DISTRIBUTE_BY,
 561        "DIV": TokenType.DIV,
 562        "DROP": TokenType.DROP,
 563        "ELSE": TokenType.ELSE,
 564        "END": TokenType.END,
 565        "ESCAPE": TokenType.ESCAPE,
 566        "EXCEPT": TokenType.EXCEPT,
 567        "EXECUTE": TokenType.EXECUTE,
 568        "EXISTS": TokenType.EXISTS,
 569        "FALSE": TokenType.FALSE,
 570        "FETCH": TokenType.FETCH,
 571        "FILTER": TokenType.FILTER,
 572        "FIRST": TokenType.FIRST,
 573        "FULL": TokenType.FULL,
 574        "FUNCTION": TokenType.FUNCTION,
 575        "FOR": TokenType.FOR,
 576        "FOREIGN KEY": TokenType.FOREIGN_KEY,
 577        "FORMAT": TokenType.FORMAT,
 578        "FROM": TokenType.FROM,
 579        "GEOGRAPHY": TokenType.GEOGRAPHY,
 580        "GEOMETRY": TokenType.GEOMETRY,
 581        "GLOB": TokenType.GLOB,
 582        "GROUP BY": TokenType.GROUP_BY,
 583        "GROUPING SETS": TokenType.GROUPING_SETS,
 584        "HAVING": TokenType.HAVING,
 585        "ILIKE": TokenType.ILIKE,
 586        "IN": TokenType.IN,
 587        "INDEX": TokenType.INDEX,
 588        "INET": TokenType.INET,
 589        "INNER": TokenType.INNER,
 590        "INSERT": TokenType.INSERT,
 591        "INTERVAL": TokenType.INTERVAL,
 592        "INTERSECT": TokenType.INTERSECT,
 593        "INTO": TokenType.INTO,
 594        "IS": TokenType.IS,
 595        "ISNULL": TokenType.ISNULL,
 596        "JOIN": TokenType.JOIN,
 597        "KEEP": TokenType.KEEP,
 598        "LATERAL": TokenType.LATERAL,
 599        "LEFT": TokenType.LEFT,
 600        "LIKE": TokenType.LIKE,
 601        "LIMIT": TokenType.LIMIT,
 602        "LOAD": TokenType.LOAD,
 603        "LOCK": TokenType.LOCK,
 604        "MERGE": TokenType.MERGE,
 605        "NATURAL": TokenType.NATURAL,
 606        "NEXT": TokenType.NEXT,
 607        "NOT": TokenType.NOT,
 608        "NOTNULL": TokenType.NOTNULL,
 609        "NULL": TokenType.NULL,
 610        "OBJECT": TokenType.OBJECT,
 611        "OFFSET": TokenType.OFFSET,
 612        "ON": TokenType.ON,
 613        "OR": TokenType.OR,
 614        "XOR": TokenType.XOR,
 615        "ORDER BY": TokenType.ORDER_BY,
 616        "ORDINALITY": TokenType.ORDINALITY,
 617        "OUTER": TokenType.OUTER,
 618        "OVER": TokenType.OVER,
 619        "OVERLAPS": TokenType.OVERLAPS,
 620        "OVERWRITE": TokenType.OVERWRITE,
 621        "PARTITION": TokenType.PARTITION,
 622        "PARTITION BY": TokenType.PARTITION_BY,
 623        "PARTITIONED BY": TokenType.PARTITION_BY,
 624        "PARTITIONED_BY": TokenType.PARTITION_BY,
 625        "PERCENT": TokenType.PERCENT,
 626        "PIVOT": TokenType.PIVOT,
 627        "PRAGMA": TokenType.PRAGMA,
 628        "PRIMARY KEY": TokenType.PRIMARY_KEY,
 629        "PROCEDURE": TokenType.PROCEDURE,
 630        "QUALIFY": TokenType.QUALIFY,
 631        "RANGE": TokenType.RANGE,
 632        "RECURSIVE": TokenType.RECURSIVE,
 633        "REGEXP": TokenType.RLIKE,
 634        "REPLACE": TokenType.REPLACE,
 635        "RETURNING": TokenType.RETURNING,
 636        "REFERENCES": TokenType.REFERENCES,
 637        "RIGHT": TokenType.RIGHT,
 638        "RLIKE": TokenType.RLIKE,
 639        "ROLLBACK": TokenType.ROLLBACK,
 640        "ROLLUP": TokenType.ROLLUP,
 641        "ROW": TokenType.ROW,
 642        "ROWS": TokenType.ROWS,
 643        "SCHEMA": TokenType.SCHEMA,
 644        "SELECT": TokenType.SELECT,
 645        "SEMI": TokenType.SEMI,
 646        "SET": TokenType.SET,
 647        "SETTINGS": TokenType.SETTINGS,
 648        "SHOW": TokenType.SHOW,
 649        "SIMILAR TO": TokenType.SIMILAR_TO,
 650        "SOME": TokenType.SOME,
 651        "SORT BY": TokenType.SORT_BY,
 652        "START WITH": TokenType.START_WITH,
 653        "TABLE": TokenType.TABLE,
 654        "TABLESAMPLE": TokenType.TABLE_SAMPLE,
 655        "TEMP": TokenType.TEMPORARY,
 656        "TEMPORARY": TokenType.TEMPORARY,
 657        "THEN": TokenType.THEN,
 658        "TRUE": TokenType.TRUE,
 659        "UNION": TokenType.UNION,
 660        "UNKNOWN": TokenType.UNKNOWN,
 661        "UNNEST": TokenType.UNNEST,
 662        "UNPIVOT": TokenType.UNPIVOT,
 663        "UPDATE": TokenType.UPDATE,
 664        "USE": TokenType.USE,
 665        "USING": TokenType.USING,
 666        "UUID": TokenType.UUID,
 667        "VALUES": TokenType.VALUES,
 668        "VIEW": TokenType.VIEW,
 669        "VOLATILE": TokenType.VOLATILE,
 670        "WHEN": TokenType.WHEN,
 671        "WHERE": TokenType.WHERE,
 672        "WINDOW": TokenType.WINDOW,
 673        "WITH": TokenType.WITH,
 674        "APPLY": TokenType.APPLY,
 675        "ARRAY": TokenType.ARRAY,
 676        "BIT": TokenType.BIT,
 677        "BOOL": TokenType.BOOLEAN,
 678        "BOOLEAN": TokenType.BOOLEAN,
 679        "BYTE": TokenType.TINYINT,
 680        "MEDIUMINT": TokenType.MEDIUMINT,
 681        "TINYINT": TokenType.TINYINT,
 682        "SHORT": TokenType.SMALLINT,
 683        "SMALLINT": TokenType.SMALLINT,
 684        "INT128": TokenType.INT128,
 685        "INT2": TokenType.SMALLINT,
 686        "INTEGER": TokenType.INT,
 687        "INT": TokenType.INT,
 688        "INT4": TokenType.INT,
 689        "LONG": TokenType.BIGINT,
 690        "BIGINT": TokenType.BIGINT,
 691        "INT8": TokenType.BIGINT,
 692        "DEC": TokenType.DECIMAL,
 693        "DECIMAL": TokenType.DECIMAL,
 694        "BIGDECIMAL": TokenType.BIGDECIMAL,
 695        "BIGNUMERIC": TokenType.BIGDECIMAL,
 696        "MAP": TokenType.MAP,
 697        "NULLABLE": TokenType.NULLABLE,
 698        "NUMBER": TokenType.DECIMAL,
 699        "NUMERIC": TokenType.DECIMAL,
 700        "FIXED": TokenType.DECIMAL,
 701        "REAL": TokenType.FLOAT,
 702        "FLOAT": TokenType.FLOAT,
 703        "FLOAT4": TokenType.FLOAT,
 704        "FLOAT8": TokenType.DOUBLE,
 705        "DOUBLE": TokenType.DOUBLE,
 706        "DOUBLE PRECISION": TokenType.DOUBLE,
 707        "JSON": TokenType.JSON,
 708        "CHAR": TokenType.CHAR,
 709        "CHARACTER": TokenType.CHAR,
 710        "NCHAR": TokenType.NCHAR,
 711        "VARCHAR": TokenType.VARCHAR,
 712        "VARCHAR2": TokenType.VARCHAR,
 713        "NVARCHAR": TokenType.NVARCHAR,
 714        "NVARCHAR2": TokenType.NVARCHAR,
 715        "STR": TokenType.TEXT,
 716        "STRING": TokenType.TEXT,
 717        "TEXT": TokenType.TEXT,
 718        "LONGTEXT": TokenType.LONGTEXT,
 719        "MEDIUMTEXT": TokenType.MEDIUMTEXT,
 720        "TINYTEXT": TokenType.TINYTEXT,
 721        "CLOB": TokenType.TEXT,
 722        "LONGVARCHAR": TokenType.TEXT,
 723        "BINARY": TokenType.BINARY,
 724        "BLOB": TokenType.VARBINARY,
 725        "LONGBLOB": TokenType.LONGBLOB,
 726        "MEDIUMBLOB": TokenType.MEDIUMBLOB,
 727        "TINYBLOB": TokenType.TINYBLOB,
 728        "BYTEA": TokenType.VARBINARY,
 729        "VARBINARY": TokenType.VARBINARY,
 730        "TIME": TokenType.TIME,
 731        "TIMETZ": TokenType.TIMETZ,
 732        "TIMESTAMP": TokenType.TIMESTAMP,
 733        "TIMESTAMPTZ": TokenType.TIMESTAMPTZ,
 734        "TIMESTAMPLTZ": TokenType.TIMESTAMPLTZ,
 735        "DATE": TokenType.DATE,
 736        "DATETIME": TokenType.DATETIME,
 737        "INT4RANGE": TokenType.INT4RANGE,
 738        "INT4MULTIRANGE": TokenType.INT4MULTIRANGE,
 739        "INT8RANGE": TokenType.INT8RANGE,
 740        "INT8MULTIRANGE": TokenType.INT8MULTIRANGE,
 741        "NUMRANGE": TokenType.NUMRANGE,
 742        "NUMMULTIRANGE": TokenType.NUMMULTIRANGE,
 743        "TSRANGE": TokenType.TSRANGE,
 744        "TSMULTIRANGE": TokenType.TSMULTIRANGE,
 745        "TSTZRANGE": TokenType.TSTZRANGE,
 746        "TSTZMULTIRANGE": TokenType.TSTZMULTIRANGE,
 747        "DATERANGE": TokenType.DATERANGE,
 748        "DATEMULTIRANGE": TokenType.DATEMULTIRANGE,
 749        "UNIQUE": TokenType.UNIQUE,
 750        "STRUCT": TokenType.STRUCT,
 751        "VARIANT": TokenType.VARIANT,
 752        "ALTER": TokenType.ALTER,
 753        "ANALYZE": TokenType.COMMAND,
 754        "CALL": TokenType.COMMAND,
 755        "COMMENT": TokenType.COMMENT,
 756        "COPY": TokenType.COMMAND,
 757        "EXPLAIN": TokenType.COMMAND,
 758        "GRANT": TokenType.COMMAND,
 759        "OPTIMIZE": TokenType.COMMAND,
 760        "PREPARE": TokenType.COMMAND,
 761        "TRUNCATE": TokenType.COMMAND,
 762        "VACUUM": TokenType.COMMAND,
 763        "USER-DEFINED": TokenType.USERDEFINED,
 764        "FOR VERSION": TokenType.VERSION_SNAPSHOT,
 765        "FOR TIMESTAMP": TokenType.TIMESTAMP_SNAPSHOT,
 766    }
 767
 768    WHITE_SPACE: t.Dict[t.Optional[str], TokenType] = {
 769        " ": TokenType.SPACE,
 770        "\t": TokenType.SPACE,
 771        "\n": TokenType.BREAK,
 772        "\r": TokenType.BREAK,
 773        "\r\n": TokenType.BREAK,
 774    }
 775
 776    COMMANDS = {
 777        TokenType.COMMAND,
 778        TokenType.EXECUTE,
 779        TokenType.FETCH,
 780        TokenType.SHOW,
 781    }
 782
 783    COMMAND_PREFIX_TOKENS = {TokenType.SEMICOLON, TokenType.BEGIN}
 784
 785    # handle numeric literals like in hive (3L = BIGINT)
 786    NUMERIC_LITERALS: t.Dict[str, str] = {}
 787    ENCODE: t.Optional[str] = None
 788
 789    COMMENTS = ["--", ("/*", "*/")]
 790
 791    __slots__ = (
 792        "sql",
 793        "size",
 794        "tokens",
 795        "_start",
 796        "_current",
 797        "_line",
 798        "_col",
 799        "_comments",
 800        "_char",
 801        "_end",
 802        "_peek",
 803        "_prev_token_line",
 804    )
 805
 806    def __init__(self) -> None:
 807        self.reset()
 808
 809    def reset(self) -> None:
 810        self.sql = ""
 811        self.size = 0
 812        self.tokens: t.List[Token] = []
 813        self._start = 0
 814        self._current = 0
 815        self._line = 1
 816        self._col = 0
 817        self._comments: t.List[str] = []
 818
 819        self._char = ""
 820        self._end = False
 821        self._peek = ""
 822        self._prev_token_line = -1
 823
 824    def tokenize(self, sql: str) -> t.List[Token]:
 825        """Returns a list of tokens corresponding to the SQL string `sql`."""
 826        self.reset()
 827        self.sql = sql
 828        self.size = len(sql)
 829
 830        try:
 831            self._scan()
 832        except Exception as e:
 833            start = max(self._current - 50, 0)
 834            end = min(self._current + 50, self.size - 1)
 835            context = self.sql[start:end]
 836            raise TokenError(f"Error tokenizing '{context}'") from e
 837
 838        return self.tokens
 839
 840    def _scan(self, until: t.Optional[t.Callable] = None) -> None:
 841        while self.size and not self._end:
 842            self._start = self._current
 843            self._advance()
 844
 845            if self._char is None:
 846                break
 847
 848            if self._char not in self.WHITE_SPACE:
 849                if self._char.isdigit():
 850                    self._scan_number()
 851                elif self._char in self._IDENTIFIERS:
 852                    self._scan_identifier(self._IDENTIFIERS[self._char])
 853                else:
 854                    self._scan_keywords()
 855
 856            if until and until():
 857                break
 858
 859        if self.tokens and self._comments:
 860            self.tokens[-1].comments.extend(self._comments)
 861
 862    def _chars(self, size: int) -> str:
 863        if size == 1:
 864            return self._char
 865
 866        start = self._current - 1
 867        end = start + size
 868
 869        return self.sql[start:end] if end <= self.size else ""
 870
 871    def _advance(self, i: int = 1, alnum: bool = False) -> None:
 872        if self.WHITE_SPACE.get(self._char) is TokenType.BREAK:
 873            self._col = 1
 874            self._line += 1
 875        else:
 876            self._col += i
 877
 878        self._current += i
 879        self._end = self._current >= self.size
 880        self._char = self.sql[self._current - 1]
 881        self._peek = "" if self._end else self.sql[self._current]
 882
 883        if alnum and self._char.isalnum():
 884            # Here we use local variables instead of attributes for better performance
 885            _col = self._col
 886            _current = self._current
 887            _end = self._end
 888            _peek = self._peek
 889
 890            while _peek.isalnum():
 891                _col += 1
 892                _current += 1
 893                _end = _current >= self.size
 894                _peek = "" if _end else self.sql[_current]
 895
 896            self._col = _col
 897            self._current = _current
 898            self._end = _end
 899            self._peek = _peek
 900            self._char = self.sql[_current - 1]
 901
 902    @property
 903    def _text(self) -> str:
 904        return self.sql[self._start : self._current]
 905
 906    def peek(self, i: int = 0) -> str:
 907        i = self._current + i
 908        if i < self.size:
 909            return self.sql[i]
 910        return ""
 911
 912    def _add(self, token_type: TokenType, text: t.Optional[str] = None) -> None:
 913        self._prev_token_line = self._line
 914
 915        if self._comments and token_type == TokenType.SEMICOLON and self.tokens:
 916            self.tokens[-1].comments.extend(self._comments)
 917            self._comments = []
 918
 919        self.tokens.append(
 920            Token(
 921                token_type,
 922                text=self._text if text is None else text,
 923                line=self._line,
 924                col=self._col,
 925                start=self._start,
 926                end=self._current - 1,
 927                comments=self._comments,
 928            )
 929        )
 930        self._comments = []
 931
 932        # If we have either a semicolon or a begin token before the command's token, we'll parse
 933        # whatever follows the command's token as a string
 934        if (
 935            token_type in self.COMMANDS
 936            and self._peek != ";"
 937            and (len(self.tokens) == 1 or self.tokens[-2].token_type in self.COMMAND_PREFIX_TOKENS)
 938        ):
 939            start = self._current
 940            tokens = len(self.tokens)
 941            self._scan(lambda: self._peek == ";")
 942            self.tokens = self.tokens[:tokens]
 943            text = self.sql[start : self._current].strip()
 944            if text:
 945                self._add(TokenType.STRING, text)
 946
 947    def _scan_keywords(self) -> None:
 948        size = 0
 949        word = None
 950        chars = self._text
 951        char = chars
 952        prev_space = False
 953        skip = False
 954        trie = self._KEYWORD_TRIE
 955        single_token = char in self.SINGLE_TOKENS
 956
 957        while chars:
 958            if skip:
 959                result = TrieResult.PREFIX
 960            else:
 961                result, trie = in_trie(trie, char.upper())
 962
 963            if result == TrieResult.FAILED:
 964                break
 965            if result == TrieResult.EXISTS:
 966                word = chars
 967
 968            end = self._current + size
 969            size += 1
 970
 971            if end < self.size:
 972                char = self.sql[end]
 973                single_token = single_token or char in self.SINGLE_TOKENS
 974                is_space = char in self.WHITE_SPACE
 975
 976                if not is_space or not prev_space:
 977                    if is_space:
 978                        char = " "
 979                    chars += char
 980                    prev_space = is_space
 981                    skip = False
 982                else:
 983                    skip = True
 984            else:
 985                char = ""
 986                chars = " "
 987
 988        if word:
 989            if self._scan_string(word):
 990                return
 991            if self._scan_comment(word):
 992                return
 993            if prev_space or single_token or not char:
 994                self._advance(size - 1)
 995                word = word.upper()
 996                self._add(self.KEYWORDS[word], text=word)
 997                return
 998        if self._char in self.SINGLE_TOKENS:
 999            self._add(self.SINGLE_TOKENS[self._char], text=self._char)
1000            return
1001        self._scan_var()
1002
1003    def _scan_comment(self, comment_start: str) -> bool:
1004        if comment_start not in self._COMMENTS:
1005            return False
1006
1007        comment_start_line = self._line
1008        comment_start_size = len(comment_start)
1009        comment_end = self._COMMENTS[comment_start]
1010
1011        if comment_end:
1012            # Skip the comment's start delimiter
1013            self._advance(comment_start_size)
1014
1015            comment_end_size = len(comment_end)
1016            while not self._end and self._chars(comment_end_size) != comment_end:
1017                self._advance(alnum=True)
1018
1019            self._comments.append(self._text[comment_start_size : -comment_end_size + 1])
1020            self._advance(comment_end_size - 1)
1021        else:
1022            while not self._end and not self.WHITE_SPACE.get(self._peek) is TokenType.BREAK:
1023                self._advance(alnum=True)
1024            self._comments.append(self._text[comment_start_size:])
1025
1026        # Leading comment is attached to the succeeding token, whilst trailing comment to the preceding.
1027        # Multiple consecutive comments are preserved by appending them to the current comments list.
1028        if comment_start_line == self._prev_token_line:
1029            self.tokens[-1].comments.extend(self._comments)
1030            self._comments = []
1031            self._prev_token_line = self._line
1032
1033        return True
1034
1035    def _scan_number(self) -> None:
1036        if self._char == "0":
1037            peek = self._peek.upper()
1038            if peek == "B":
1039                return self._scan_bits() if self.BIT_STRINGS else self._add(TokenType.NUMBER)
1040            elif peek == "X":
1041                return self._scan_hex() if self.HEX_STRINGS else self._add(TokenType.NUMBER)
1042
1043        decimal = False
1044        scientific = 0
1045
1046        while True:
1047            if self._peek.isdigit():
1048                self._advance()
1049            elif self._peek == "." and not decimal:
1050                after = self.peek(1)
1051                if after.isdigit() or not after.isalpha():
1052                    decimal = True
1053                    self._advance()
1054                else:
1055                    return self._add(TokenType.VAR)
1056            elif self._peek in ("-", "+") and scientific == 1:
1057                scientific += 1
1058                self._advance()
1059            elif self._peek.upper() == "E" and not scientific:
1060                scientific += 1
1061                self._advance()
1062            elif self._peek.isidentifier():
1063                number_text = self._text
1064                literal = ""
1065
1066                while self._peek.strip() and self._peek not in self.SINGLE_TOKENS:
1067                    literal += self._peek.upper()
1068                    self._advance()
1069
1070                token_type = self.KEYWORDS.get(self.NUMERIC_LITERALS.get(literal, ""))
1071
1072                if token_type:
1073                    self._add(TokenType.NUMBER, number_text)
1074                    self._add(TokenType.DCOLON, "::")
1075                    return self._add(token_type, literal)
1076                elif self.IDENTIFIERS_CAN_START_WITH_DIGIT:
1077                    return self._add(TokenType.VAR)
1078
1079                self._advance(-len(literal))
1080                return self._add(TokenType.NUMBER, number_text)
1081            else:
1082                return self._add(TokenType.NUMBER)
1083
1084    def _scan_bits(self) -> None:
1085        self._advance()
1086        value = self._extract_value()
1087        try:
1088            # If `value` can't be converted to a binary, fallback to tokenizing it as an identifier
1089            int(value, 2)
1090            self._add(TokenType.BIT_STRING, value[2:])  # Drop the 0b
1091        except ValueError:
1092            self._add(TokenType.IDENTIFIER)
1093
1094    def _scan_hex(self) -> None:
1095        self._advance()
1096        value = self._extract_value()
1097        try:
1098            # If `value` can't be converted to a hex, fallback to tokenizing it as an identifier
1099            int(value, 16)
1100            self._add(TokenType.HEX_STRING, value[2:])  # Drop the 0x
1101        except ValueError:
1102            self._add(TokenType.IDENTIFIER)
1103
1104    def _extract_value(self) -> str:
1105        while True:
1106            char = self._peek.strip()
1107            if char and char not in self.SINGLE_TOKENS:
1108                self._advance(alnum=True)
1109            else:
1110                break
1111
1112        return self._text
1113
1114    def _scan_string(self, start: str) -> bool:
1115        base = None
1116        token_type = TokenType.STRING
1117
1118        if start in self._QUOTES:
1119            end = self._QUOTES[start]
1120        elif start in self._FORMAT_STRINGS:
1121            end, token_type = self._FORMAT_STRINGS[start]
1122
1123            if token_type == TokenType.HEX_STRING:
1124                base = 16
1125            elif token_type == TokenType.BIT_STRING:
1126                base = 2
1127        else:
1128            return False
1129
1130        self._advance(len(start))
1131        text = self._extract_string(end)
1132
1133        if base:
1134            try:
1135                int(text, base)
1136            except:
1137                raise TokenError(
1138                    f"Numeric string contains invalid characters from {self._line}:{self._start}"
1139                )
1140        else:
1141            text = text.encode(self.ENCODE).decode(self.ENCODE) if self.ENCODE else text
1142
1143        self._add(token_type, text)
1144        return True
1145
1146    def _scan_identifier(self, identifier_end: str) -> None:
1147        self._advance()
1148        text = self._extract_string(identifier_end, self._IDENTIFIER_ESCAPES)
1149        self._add(TokenType.IDENTIFIER, text)
1150
1151    def _scan_var(self) -> None:
1152        while True:
1153            char = self._peek.strip()
1154            if char and (char in self.VAR_SINGLE_TOKENS or char not in self.SINGLE_TOKENS):
1155                self._advance(alnum=True)
1156            else:
1157                break
1158
1159        self._add(
1160            TokenType.VAR
1161            if self.tokens and self.tokens[-1].token_type == TokenType.PARAMETER
1162            else self.KEYWORDS.get(self._text.upper(), TokenType.VAR)
1163        )
1164
1165    def _extract_string(self, delimiter: str, escapes=None) -> str:
1166        text = ""
1167        delim_size = len(delimiter)
1168        escapes = self._STRING_ESCAPES if escapes is None else escapes
1169
1170        while True:
1171            if self._char in escapes and (self._peek == delimiter or self._peek in escapes):
1172                if self._peek == delimiter:
1173                    text += self._peek
1174                else:
1175                    text += self._char + self._peek
1176
1177                if self._current + 1 < self.size:
1178                    self._advance(2)
1179                else:
1180                    raise TokenError(f"Missing {delimiter} from {self._line}:{self._current}")
1181            else:
1182                if self._chars(delim_size) == delimiter:
1183                    if delim_size > 1:
1184                        self._advance(delim_size - 1)
1185                    break
1186
1187                if self._end:
1188                    raise TokenError(f"Missing {delimiter} from {self._line}:{self._start}")
1189
1190                current = self._current - 1
1191                self._advance(alnum=True)
1192                text += self.sql[current : self._current - 1]
1193
1194        return text