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