sqlglot.tokens

   1from __future__ import annotations
   2
   3import typing as t
   4from enum import auto
   5
   6from sqlglot.helper import AutoName
   7from sqlglot.trie import in_trie, new_trie
   8
   9
  10class TokenType(AutoName):
  11    L_PAREN = auto()
  12    R_PAREN = auto()
  13    L_BRACKET = auto()
  14    R_BRACKET = auto()
  15    L_BRACE = auto()
  16    R_BRACE = auto()
  17    COMMA = auto()
  18    DOT = auto()
  19    DASH = auto()
  20    PLUS = auto()
  21    COLON = auto()
  22    DCOLON = auto()
  23    SEMICOLON = auto()
  24    STAR = auto()
  25    BACKSLASH = auto()
  26    SLASH = auto()
  27    LT = auto()
  28    LTE = auto()
  29    GT = auto()
  30    GTE = auto()
  31    NOT = auto()
  32    EQ = auto()
  33    NEQ = auto()
  34    NULLSAFE_EQ = auto()
  35    AND = auto()
  36    OR = auto()
  37    AMP = auto()
  38    DPIPE = auto()
  39    PIPE = auto()
  40    CARET = auto()
  41    TILDA = auto()
  42    ARROW = auto()
  43    DARROW = auto()
  44    FARROW = auto()
  45    HASH = auto()
  46    HASH_ARROW = auto()
  47    DHASH_ARROW = auto()
  48    LR_ARROW = auto()
  49    DOLLAR = auto()
  50    PARAMETER = auto()
  51    SESSION_PARAMETER = auto()
  52    NATIONAL = auto()
  53
  54    BLOCK_START = auto()
  55    BLOCK_END = auto()
  56
  57    SPACE = auto()
  58    BREAK = auto()
  59
  60    STRING = auto()
  61    NUMBER = auto()
  62    IDENTIFIER = auto()
  63    COLUMN = auto()
  64    COLUMN_DEF = auto()
  65    SCHEMA = auto()
  66    TABLE = auto()
  67    VAR = auto()
  68    BIT_STRING = auto()
  69    HEX_STRING = auto()
  70    BYTE_STRING = auto()
  71
  72    # types
  73    BOOLEAN = auto()
  74    TINYINT = auto()
  75    SMALLINT = auto()
  76    INT = auto()
  77    BIGINT = auto()
  78    FLOAT = auto()
  79    DOUBLE = auto()
  80    DECIMAL = auto()
  81    CHAR = auto()
  82    NCHAR = auto()
  83    VARCHAR = auto()
  84    NVARCHAR = auto()
  85    TEXT = auto()
  86    MEDIUMTEXT = auto()
  87    LONGTEXT = auto()
  88    MEDIUMBLOB = auto()
  89    LONGBLOB = auto()
  90    BINARY = auto()
  91    VARBINARY = auto()
  92    JSON = auto()
  93    JSONB = auto()
  94    TIME = auto()
  95    TIMESTAMP = auto()
  96    TIMESTAMPTZ = auto()
  97    TIMESTAMPLTZ = auto()
  98    DATETIME = auto()
  99    DATE = auto()
 100    UUID = auto()
 101    GEOGRAPHY = auto()
 102    NULLABLE = auto()
 103    GEOMETRY = auto()
 104    HLLSKETCH = auto()
 105    HSTORE = auto()
 106    SUPER = auto()
 107    SERIAL = auto()
 108    SMALLSERIAL = auto()
 109    BIGSERIAL = auto()
 110    XML = auto()
 111    UNIQUEIDENTIFIER = auto()
 112    MONEY = auto()
 113    SMALLMONEY = auto()
 114    ROWVERSION = auto()
 115    IMAGE = auto()
 116    VARIANT = auto()
 117    OBJECT = auto()
 118
 119    # keywords
 120    ALIAS = auto()
 121    ALTER = auto()
 122    ALWAYS = auto()
 123    ALL = auto()
 124    ANTI = auto()
 125    ANY = auto()
 126    APPLY = auto()
 127    ARRAY = auto()
 128    ASC = auto()
 129    ASOF = auto()
 130    AT_TIME_ZONE = auto()
 131    AUTO_INCREMENT = auto()
 132    BEGIN = auto()
 133    BETWEEN = auto()
 134    BOTH = auto()
 135    BUCKET = auto()
 136    BY_DEFAULT = auto()
 137    CACHE = auto()
 138    CASCADE = auto()
 139    CASE = auto()
 140    CHARACTER_SET = auto()
 141    CLUSTER_BY = auto()
 142    COLLATE = auto()
 143    COMMAND = auto()
 144    COMMENT = auto()
 145    COMMIT = auto()
 146    COMPOUND = auto()
 147    CONSTRAINT = auto()
 148    CREATE = auto()
 149    CROSS = auto()
 150    CUBE = auto()
 151    CURRENT_DATE = auto()
 152    CURRENT_DATETIME = auto()
 153    CURRENT_ROW = auto()
 154    CURRENT_TIME = auto()
 155    CURRENT_TIMESTAMP = auto()
 156    DEFAULT = auto()
 157    DELETE = auto()
 158    DESC = auto()
 159    DESCRIBE = auto()
 160    DISTINCT = auto()
 161    DISTINCT_FROM = auto()
 162    DISTRIBUTE_BY = auto()
 163    DIV = auto()
 164    DROP = auto()
 165    ELSE = auto()
 166    END = auto()
 167    ESCAPE = auto()
 168    EXCEPT = auto()
 169    EXECUTE = auto()
 170    EXISTS = auto()
 171    FALSE = auto()
 172    FETCH = auto()
 173    FILTER = auto()
 174    FINAL = auto()
 175    FIRST = auto()
 176    FOLLOWING = auto()
 177    FOR = auto()
 178    FOREIGN_KEY = auto()
 179    FORMAT = auto()
 180    FROM = auto()
 181    FULL = auto()
 182    FUNCTION = auto()
 183    GLOB = auto()
 184    GLOBAL = auto()
 185    GROUP_BY = auto()
 186    GROUPING_SETS = auto()
 187    HAVING = auto()
 188    HINT = auto()
 189    IF = auto()
 190    IGNORE_NULLS = auto()
 191    ILIKE = auto()
 192    ILIKE_ANY = auto()
 193    IN = auto()
 194    INDEX = auto()
 195    INNER = auto()
 196    INSERT = auto()
 197    INTERSECT = auto()
 198    INTERVAL = auto()
 199    INTO = auto()
 200    INTRODUCER = auto()
 201    IRLIKE = auto()
 202    IS = auto()
 203    ISNULL = auto()
 204    JOIN = auto()
 205    LANGUAGE = auto()
 206    LATERAL = auto()
 207    LAZY = auto()
 208    LEADING = auto()
 209    LEFT = auto()
 210    LIKE = auto()
 211    LIKE_ANY = auto()
 212    LIMIT = auto()
 213    LOAD_DATA = auto()
 214    LOCAL = auto()
 215    MAP = auto()
 216    MATCH_RECOGNIZE = auto()
 217    MATERIALIZED = auto()
 218    MERGE = auto()
 219    MOD = auto()
 220    NATURAL = auto()
 221    NEXT = auto()
 222    NO_ACTION = auto()
 223    NOTNULL = auto()
 224    NULL = auto()
 225    NULLS_FIRST = auto()
 226    NULLS_LAST = auto()
 227    OFFSET = auto()
 228    ON = auto()
 229    ONLY = auto()
 230    OPTIONS = auto()
 231    ORDER_BY = auto()
 232    ORDERED = auto()
 233    ORDINALITY = auto()
 234    OUTER = auto()
 235    OUT_OF = auto()
 236    OVER = auto()
 237    OVERWRITE = auto()
 238    PARTITION = auto()
 239    PARTITION_BY = auto()
 240    PERCENT = auto()
 241    PIVOT = auto()
 242    PLACEHOLDER = auto()
 243    PRECEDING = auto()
 244    PRIMARY_KEY = auto()
 245    PROCEDURE = auto()
 246    PROPERTIES = auto()
 247    PSEUDO_TYPE = auto()
 248    QUALIFY = auto()
 249    QUOTE = auto()
 250    RANGE = auto()
 251    RECURSIVE = auto()
 252    REPLACE = auto()
 253    RESPECT_NULLS = auto()
 254    RETURNING = auto()
 255    REFERENCES = auto()
 256    RIGHT = auto()
 257    RLIKE = auto()
 258    ROLLBACK = auto()
 259    ROLLUP = auto()
 260    ROW = auto()
 261    ROWS = auto()
 262    SEED = auto()
 263    SELECT = auto()
 264    SEMI = auto()
 265    SEPARATOR = auto()
 266    SERDE_PROPERTIES = auto()
 267    SET = auto()
 268    SHOW = auto()
 269    SIMILAR_TO = auto()
 270    SOME = auto()
 271    SORTKEY = auto()
 272    SORT_BY = auto()
 273    STRUCT = auto()
 274    TABLE_SAMPLE = auto()
 275    TEMPORARY = auto()
 276    TOP = auto()
 277    THEN = auto()
 278    TRAILING = auto()
 279    TRUE = auto()
 280    UNBOUNDED = auto()
 281    UNCACHE = auto()
 282    UNION = auto()
 283    UNLOGGED = auto()
 284    UNNEST = auto()
 285    UNPIVOT = auto()
 286    UPDATE = auto()
 287    USE = auto()
 288    USING = auto()
 289    VALUES = auto()
 290    VIEW = auto()
 291    VOLATILE = auto()
 292    WHEN = auto()
 293    WHERE = auto()
 294    WINDOW = auto()
 295    WITH = auto()
 296    WITH_TIME_ZONE = auto()
 297    WITH_LOCAL_TIME_ZONE = auto()
 298    WITHIN_GROUP = auto()
 299    WITHOUT_TIME_ZONE = auto()
 300    UNIQUE = auto()
 301
 302
 303class Token:
 304    __slots__ = ("token_type", "text", "line", "col", "comments")
 305
 306    @classmethod
 307    def number(cls, number: int) -> Token:
 308        """Returns a NUMBER token with `number` as its text."""
 309        return cls(TokenType.NUMBER, str(number))
 310
 311    @classmethod
 312    def string(cls, string: str) -> Token:
 313        """Returns a STRING token with `string` as its text."""
 314        return cls(TokenType.STRING, string)
 315
 316    @classmethod
 317    def identifier(cls, identifier: str) -> Token:
 318        """Returns an IDENTIFIER token with `identifier` as its text."""
 319        return cls(TokenType.IDENTIFIER, identifier)
 320
 321    @classmethod
 322    def var(cls, var: str) -> Token:
 323        """Returns an VAR token with `var` as its text."""
 324        return cls(TokenType.VAR, var)
 325
 326    def __init__(
 327        self,
 328        token_type: TokenType,
 329        text: str,
 330        line: int = 1,
 331        col: int = 1,
 332        comments: t.List[str] = [],
 333    ) -> None:
 334        self.token_type = token_type
 335        self.text = text
 336        self.line = line
 337        self.col = max(col - len(text), 1)
 338        self.comments = comments
 339
 340    def __repr__(self) -> str:
 341        attributes = ", ".join(f"{k}: {getattr(self, k)}" for k in self.__slots__)
 342        return f"<Token {attributes}>"
 343
 344
 345class _Tokenizer(type):
 346    def __new__(cls, clsname, bases, attrs):  # type: ignore
 347        klass = super().__new__(cls, clsname, bases, attrs)
 348
 349        klass._QUOTES = {
 350            f"{prefix}{s}": e
 351            for s, e in cls._delimeter_list_to_dict(klass.QUOTES).items()
 352            for prefix in (("",) if s[0].isalpha() else ("", "n", "N"))
 353        }
 354        klass._BIT_STRINGS = cls._delimeter_list_to_dict(klass.BIT_STRINGS)
 355        klass._HEX_STRINGS = cls._delimeter_list_to_dict(klass.HEX_STRINGS)
 356        klass._BYTE_STRINGS = cls._delimeter_list_to_dict(klass.BYTE_STRINGS)
 357        klass._IDENTIFIERS = cls._delimeter_list_to_dict(klass.IDENTIFIERS)
 358        klass._STRING_ESCAPES = set(klass.STRING_ESCAPES)
 359        klass._IDENTIFIER_ESCAPES = set(klass.IDENTIFIER_ESCAPES)
 360        klass._COMMENTS = dict(
 361            (comment, None) if isinstance(comment, str) else (comment[0], comment[1])
 362            for comment in klass.COMMENTS
 363        )
 364
 365        klass.KEYWORD_TRIE = new_trie(
 366            key.upper()
 367            for key in {
 368                **klass.KEYWORDS,
 369                **{comment: TokenType.COMMENT for comment in klass._COMMENTS},
 370                **{quote: TokenType.QUOTE for quote in klass._QUOTES},
 371                **{bit_string: TokenType.BIT_STRING for bit_string in klass._BIT_STRINGS},
 372                **{hex_string: TokenType.HEX_STRING for hex_string in klass._HEX_STRINGS},
 373                **{byte_string: TokenType.BYTE_STRING for byte_string in klass._BYTE_STRINGS},
 374            }
 375            if " " in key or any(single in key for single in klass.SINGLE_TOKENS)
 376        )
 377
 378        return klass
 379
 380    @staticmethod
 381    def _delimeter_list_to_dict(list: t.List[str | t.Tuple[str, str]]) -> t.Dict[str, str]:
 382        return dict((item, item) if isinstance(item, str) else (item[0], item[1]) for item in list)
 383
 384
 385class Tokenizer(metaclass=_Tokenizer):
 386    SINGLE_TOKENS = {
 387        "(": TokenType.L_PAREN,
 388        ")": TokenType.R_PAREN,
 389        "[": TokenType.L_BRACKET,
 390        "]": TokenType.R_BRACKET,
 391        "{": TokenType.L_BRACE,
 392        "}": TokenType.R_BRACE,
 393        "&": TokenType.AMP,
 394        "^": TokenType.CARET,
 395        ":": TokenType.COLON,
 396        ",": TokenType.COMMA,
 397        ".": TokenType.DOT,
 398        "-": TokenType.DASH,
 399        "=": TokenType.EQ,
 400        ">": TokenType.GT,
 401        "<": TokenType.LT,
 402        "%": TokenType.MOD,
 403        "!": TokenType.NOT,
 404        "|": TokenType.PIPE,
 405        "+": TokenType.PLUS,
 406        ";": TokenType.SEMICOLON,
 407        "/": TokenType.SLASH,
 408        "\\": TokenType.BACKSLASH,
 409        "*": TokenType.STAR,
 410        "~": TokenType.TILDA,
 411        "?": TokenType.PLACEHOLDER,
 412        "@": TokenType.PARAMETER,
 413        # used for breaking a var like x'y' but nothing else
 414        # the token type doesn't matter
 415        "'": TokenType.QUOTE,
 416        "`": TokenType.IDENTIFIER,
 417        '"': TokenType.IDENTIFIER,
 418        "#": TokenType.HASH,
 419    }
 420
 421    QUOTES: t.List[t.Tuple[str, str] | str] = ["'"]
 422
 423    BIT_STRINGS: t.List[str | t.Tuple[str, str]] = []
 424
 425    HEX_STRINGS: t.List[str | t.Tuple[str, str]] = []
 426
 427    BYTE_STRINGS: t.List[str | t.Tuple[str, str]] = []
 428
 429    IDENTIFIERS: t.List[str | t.Tuple[str, str]] = ['"']
 430
 431    STRING_ESCAPES = ["'"]
 432
 433    _STRING_ESCAPES: t.Set[str] = set()
 434
 435    IDENTIFIER_ESCAPES = ['"']
 436
 437    _IDENTIFIER_ESCAPES: t.Set[str] = set()
 438
 439    KEYWORDS = {
 440        **{
 441            f"{key}{postfix}": TokenType.BLOCK_START
 442            for key in ("{%", "{#")
 443            for postfix in ("", "+", "-")
 444        },
 445        **{
 446            f"{prefix}{key}": TokenType.BLOCK_END
 447            for key in ("%}", "#}")
 448            for prefix in ("", "+", "-")
 449        },
 450        "{{+": TokenType.BLOCK_START,
 451        "{{-": TokenType.BLOCK_START,
 452        "+}}": TokenType.BLOCK_END,
 453        "-}}": TokenType.BLOCK_END,
 454        "/*+": TokenType.HINT,
 455        "==": TokenType.EQ,
 456        "::": TokenType.DCOLON,
 457        "||": TokenType.DPIPE,
 458        ">=": TokenType.GTE,
 459        "<=": TokenType.LTE,
 460        "<>": TokenType.NEQ,
 461        "!=": TokenType.NEQ,
 462        "<=>": TokenType.NULLSAFE_EQ,
 463        "->": TokenType.ARROW,
 464        "->>": TokenType.DARROW,
 465        "=>": TokenType.FARROW,
 466        "#>": TokenType.HASH_ARROW,
 467        "#>>": TokenType.DHASH_ARROW,
 468        "<->": TokenType.LR_ARROW,
 469        "ALL": TokenType.ALL,
 470        "ALWAYS": TokenType.ALWAYS,
 471        "AND": TokenType.AND,
 472        "ANTI": TokenType.ANTI,
 473        "ANY": TokenType.ANY,
 474        "ASC": TokenType.ASC,
 475        "AS": TokenType.ALIAS,
 476        "AT TIME ZONE": TokenType.AT_TIME_ZONE,
 477        "AUTOINCREMENT": TokenType.AUTO_INCREMENT,
 478        "AUTO_INCREMENT": TokenType.AUTO_INCREMENT,
 479        "BEGIN": TokenType.BEGIN,
 480        "BETWEEN": TokenType.BETWEEN,
 481        "BOTH": TokenType.BOTH,
 482        "BUCKET": TokenType.BUCKET,
 483        "BY DEFAULT": TokenType.BY_DEFAULT,
 484        "CACHE": TokenType.CACHE,
 485        "UNCACHE": TokenType.UNCACHE,
 486        "CASE": TokenType.CASE,
 487        "CASCADE": TokenType.CASCADE,
 488        "CHARACTER SET": TokenType.CHARACTER_SET,
 489        "CLUSTER BY": TokenType.CLUSTER_BY,
 490        "COLLATE": TokenType.COLLATE,
 491        "COLUMN": TokenType.COLUMN,
 492        "COMMIT": TokenType.COMMIT,
 493        "COMPOUND": TokenType.COMPOUND,
 494        "CONSTRAINT": TokenType.CONSTRAINT,
 495        "CREATE": TokenType.CREATE,
 496        "CROSS": TokenType.CROSS,
 497        "CUBE": TokenType.CUBE,
 498        "CURRENT_DATE": TokenType.CURRENT_DATE,
 499        "CURRENT ROW": TokenType.CURRENT_ROW,
 500        "CURRENT_TIMESTAMP": TokenType.CURRENT_TIMESTAMP,
 501        "DEFAULT": TokenType.DEFAULT,
 502        "DELETE": TokenType.DELETE,
 503        "DESC": TokenType.DESC,
 504        "DESCRIBE": TokenType.DESCRIBE,
 505        "DISTINCT": TokenType.DISTINCT,
 506        "DISTINCT FROM": TokenType.DISTINCT_FROM,
 507        "DISTRIBUTE BY": TokenType.DISTRIBUTE_BY,
 508        "DIV": TokenType.DIV,
 509        "DROP": TokenType.DROP,
 510        "ELSE": TokenType.ELSE,
 511        "END": TokenType.END,
 512        "ESCAPE": TokenType.ESCAPE,
 513        "EXCEPT": TokenType.EXCEPT,
 514        "EXECUTE": TokenType.EXECUTE,
 515        "EXISTS": TokenType.EXISTS,
 516        "FALSE": TokenType.FALSE,
 517        "FETCH": TokenType.FETCH,
 518        "FILTER": TokenType.FILTER,
 519        "FIRST": TokenType.FIRST,
 520        "FULL": TokenType.FULL,
 521        "FUNCTION": TokenType.FUNCTION,
 522        "FOLLOWING": TokenType.FOLLOWING,
 523        "FOR": TokenType.FOR,
 524        "FOREIGN KEY": TokenType.FOREIGN_KEY,
 525        "FORMAT": TokenType.FORMAT,
 526        "FROM": TokenType.FROM,
 527        "GLOB": TokenType.GLOB,
 528        "GROUP BY": TokenType.GROUP_BY,
 529        "GROUPING SETS": TokenType.GROUPING_SETS,
 530        "HAVING": TokenType.HAVING,
 531        "IF": TokenType.IF,
 532        "ILIKE": TokenType.ILIKE,
 533        "IGNORE NULLS": TokenType.IGNORE_NULLS,
 534        "IN": TokenType.IN,
 535        "INDEX": TokenType.INDEX,
 536        "INNER": TokenType.INNER,
 537        "INSERT": TokenType.INSERT,
 538        "INTERVAL": TokenType.INTERVAL,
 539        "INTERSECT": TokenType.INTERSECT,
 540        "INTO": TokenType.INTO,
 541        "IS": TokenType.IS,
 542        "ISNULL": TokenType.ISNULL,
 543        "JOIN": TokenType.JOIN,
 544        "LATERAL": TokenType.LATERAL,
 545        "LAZY": TokenType.LAZY,
 546        "LEADING": TokenType.LEADING,
 547        "LEFT": TokenType.LEFT,
 548        "LIKE": TokenType.LIKE,
 549        "LIMIT": TokenType.LIMIT,
 550        "LOAD DATA": TokenType.LOAD_DATA,
 551        "LOCAL": TokenType.LOCAL,
 552        "MATERIALIZED": TokenType.MATERIALIZED,
 553        "MERGE": TokenType.MERGE,
 554        "NATURAL": TokenType.NATURAL,
 555        "NEXT": TokenType.NEXT,
 556        "NO ACTION": TokenType.NO_ACTION,
 557        "NOT": TokenType.NOT,
 558        "NOTNULL": TokenType.NOTNULL,
 559        "NULL": TokenType.NULL,
 560        "NULLS FIRST": TokenType.NULLS_FIRST,
 561        "NULLS LAST": TokenType.NULLS_LAST,
 562        "OBJECT": TokenType.OBJECT,
 563        "OFFSET": TokenType.OFFSET,
 564        "ON": TokenType.ON,
 565        "ONLY": TokenType.ONLY,
 566        "OPTIONS": TokenType.OPTIONS,
 567        "OR": TokenType.OR,
 568        "ORDER BY": TokenType.ORDER_BY,
 569        "ORDINALITY": TokenType.ORDINALITY,
 570        "OUTER": TokenType.OUTER,
 571        "OUT OF": TokenType.OUT_OF,
 572        "OVER": TokenType.OVER,
 573        "OVERWRITE": TokenType.OVERWRITE,
 574        "PARTITION": TokenType.PARTITION,
 575        "PARTITION BY": TokenType.PARTITION_BY,
 576        "PARTITIONED BY": TokenType.PARTITION_BY,
 577        "PARTITIONED_BY": TokenType.PARTITION_BY,
 578        "PERCENT": TokenType.PERCENT,
 579        "PIVOT": TokenType.PIVOT,
 580        "PRECEDING": TokenType.PRECEDING,
 581        "PRIMARY KEY": TokenType.PRIMARY_KEY,
 582        "PROCEDURE": TokenType.PROCEDURE,
 583        "QUALIFY": TokenType.QUALIFY,
 584        "RANGE": TokenType.RANGE,
 585        "RECURSIVE": TokenType.RECURSIVE,
 586        "REGEXP": TokenType.RLIKE,
 587        "REPLACE": TokenType.REPLACE,
 588        "RESPECT NULLS": TokenType.RESPECT_NULLS,
 589        "REFERENCES": TokenType.REFERENCES,
 590        "RIGHT": TokenType.RIGHT,
 591        "RLIKE": TokenType.RLIKE,
 592        "ROLLBACK": TokenType.ROLLBACK,
 593        "ROLLUP": TokenType.ROLLUP,
 594        "ROW": TokenType.ROW,
 595        "ROWS": TokenType.ROWS,
 596        "SCHEMA": TokenType.SCHEMA,
 597        "SEED": TokenType.SEED,
 598        "SELECT": TokenType.SELECT,
 599        "SEMI": TokenType.SEMI,
 600        "SET": TokenType.SET,
 601        "SHOW": TokenType.SHOW,
 602        "SIMILAR TO": TokenType.SIMILAR_TO,
 603        "SOME": TokenType.SOME,
 604        "SORTKEY": TokenType.SORTKEY,
 605        "SORT BY": TokenType.SORT_BY,
 606        "TABLE": TokenType.TABLE,
 607        "TABLESAMPLE": TokenType.TABLE_SAMPLE,
 608        "TEMP": TokenType.TEMPORARY,
 609        "TEMPORARY": TokenType.TEMPORARY,
 610        "THEN": TokenType.THEN,
 611        "TRUE": TokenType.TRUE,
 612        "TRAILING": TokenType.TRAILING,
 613        "UNBOUNDED": TokenType.UNBOUNDED,
 614        "UNION": TokenType.UNION,
 615        "UNLOGGED": TokenType.UNLOGGED,
 616        "UNNEST": TokenType.UNNEST,
 617        "UNPIVOT": TokenType.UNPIVOT,
 618        "UPDATE": TokenType.UPDATE,
 619        "USE": TokenType.USE,
 620        "USING": TokenType.USING,
 621        "VALUES": TokenType.VALUES,
 622        "VIEW": TokenType.VIEW,
 623        "VOLATILE": TokenType.VOLATILE,
 624        "WHEN": TokenType.WHEN,
 625        "WHERE": TokenType.WHERE,
 626        "WINDOW": TokenType.WINDOW,
 627        "WITH": TokenType.WITH,
 628        "WITH TIME ZONE": TokenType.WITH_TIME_ZONE,
 629        "WITH LOCAL TIME ZONE": TokenType.WITH_LOCAL_TIME_ZONE,
 630        "WITHIN GROUP": TokenType.WITHIN_GROUP,
 631        "WITHOUT TIME ZONE": TokenType.WITHOUT_TIME_ZONE,
 632        "APPLY": TokenType.APPLY,
 633        "ARRAY": TokenType.ARRAY,
 634        "BOOL": TokenType.BOOLEAN,
 635        "BOOLEAN": TokenType.BOOLEAN,
 636        "BYTE": TokenType.TINYINT,
 637        "TINYINT": TokenType.TINYINT,
 638        "SHORT": TokenType.SMALLINT,
 639        "SMALLINT": TokenType.SMALLINT,
 640        "INT2": TokenType.SMALLINT,
 641        "INTEGER": TokenType.INT,
 642        "INT": TokenType.INT,
 643        "INT4": TokenType.INT,
 644        "LONG": TokenType.BIGINT,
 645        "BIGINT": TokenType.BIGINT,
 646        "INT8": TokenType.BIGINT,
 647        "DECIMAL": TokenType.DECIMAL,
 648        "MAP": TokenType.MAP,
 649        "NULLABLE": TokenType.NULLABLE,
 650        "NUMBER": TokenType.DECIMAL,
 651        "NUMERIC": TokenType.DECIMAL,
 652        "FIXED": TokenType.DECIMAL,
 653        "REAL": TokenType.FLOAT,
 654        "FLOAT": TokenType.FLOAT,
 655        "FLOAT4": TokenType.FLOAT,
 656        "FLOAT8": TokenType.DOUBLE,
 657        "DOUBLE": TokenType.DOUBLE,
 658        "DOUBLE PRECISION": TokenType.DOUBLE,
 659        "JSON": TokenType.JSON,
 660        "CHAR": TokenType.CHAR,
 661        "NCHAR": TokenType.NCHAR,
 662        "VARCHAR": TokenType.VARCHAR,
 663        "VARCHAR2": TokenType.VARCHAR,
 664        "NVARCHAR": TokenType.NVARCHAR,
 665        "NVARCHAR2": TokenType.NVARCHAR,
 666        "STR": TokenType.TEXT,
 667        "STRING": TokenType.TEXT,
 668        "TEXT": TokenType.TEXT,
 669        "CLOB": TokenType.TEXT,
 670        "LONGVARCHAR": TokenType.TEXT,
 671        "BINARY": TokenType.BINARY,
 672        "BLOB": TokenType.VARBINARY,
 673        "BYTEA": TokenType.VARBINARY,
 674        "VARBINARY": TokenType.VARBINARY,
 675        "TIME": TokenType.TIME,
 676        "TIMESTAMP": TokenType.TIMESTAMP,
 677        "TIMESTAMPTZ": TokenType.TIMESTAMPTZ,
 678        "TIMESTAMPLTZ": TokenType.TIMESTAMPLTZ,
 679        "DATE": TokenType.DATE,
 680        "DATETIME": TokenType.DATETIME,
 681        "UNIQUE": TokenType.UNIQUE,
 682        "STRUCT": TokenType.STRUCT,
 683        "VARIANT": TokenType.VARIANT,
 684        "ALTER": TokenType.ALTER,
 685        "ALTER AGGREGATE": TokenType.COMMAND,
 686        "ALTER DEFAULT": TokenType.COMMAND,
 687        "ALTER DOMAIN": TokenType.COMMAND,
 688        "ALTER ROLE": TokenType.COMMAND,
 689        "ALTER RULE": TokenType.COMMAND,
 690        "ALTER SEQUENCE": TokenType.COMMAND,
 691        "ALTER TYPE": TokenType.COMMAND,
 692        "ALTER USER": TokenType.COMMAND,
 693        "ALTER VIEW": TokenType.COMMAND,
 694        "ANALYZE": TokenType.COMMAND,
 695        "CALL": TokenType.COMMAND,
 696        "COPY": TokenType.COMMAND,
 697        "EXPLAIN": TokenType.COMMAND,
 698        "OPTIMIZE": TokenType.COMMAND,
 699        "PREPARE": TokenType.COMMAND,
 700        "TRUNCATE": TokenType.COMMAND,
 701        "VACUUM": TokenType.COMMAND,
 702    }
 703
 704    WHITE_SPACE = {
 705        " ": TokenType.SPACE,
 706        "\t": TokenType.SPACE,
 707        "\n": TokenType.BREAK,
 708        "\r": TokenType.BREAK,
 709        "\r\n": TokenType.BREAK,
 710    }
 711
 712    COMMANDS = {
 713        TokenType.COMMAND,
 714        TokenType.EXECUTE,
 715        TokenType.FETCH,
 716        TokenType.SET,
 717        TokenType.SHOW,
 718    }
 719
 720    COMMAND_PREFIX_TOKENS = {TokenType.SEMICOLON, TokenType.BEGIN}
 721
 722    # handle numeric literals like in hive (3L = BIGINT)
 723    NUMERIC_LITERALS: t.Dict[str, str] = {}
 724    ENCODE: t.Optional[str] = None
 725
 726    COMMENTS = ["--", ("/*", "*/")]
 727    KEYWORD_TRIE = None  # autofilled
 728
 729    IDENTIFIER_CAN_START_WITH_DIGIT = False
 730
 731    __slots__ = (
 732        "sql",
 733        "size",
 734        "tokens",
 735        "_start",
 736        "_current",
 737        "_line",
 738        "_col",
 739        "_comments",
 740        "_char",
 741        "_end",
 742        "_peek",
 743        "_prev_token_line",
 744        "_prev_token_comments",
 745        "_prev_token_type",
 746    )
 747
 748    def __init__(self) -> None:
 749        self.reset()
 750
 751    def reset(self) -> None:
 752        self.sql = ""
 753        self.size = 0
 754        self.tokens: t.List[Token] = []
 755        self._start = 0
 756        self._current = 0
 757        self._line = 1
 758        self._col = 1
 759        self._comments: t.List[str] = []
 760
 761        self._char = None
 762        self._end = None
 763        self._peek = None
 764        self._prev_token_line = -1
 765        self._prev_token_comments: t.List[str] = []
 766        self._prev_token_type = None
 767
 768    def tokenize(self, sql: str) -> t.List[Token]:
 769        """Returns a list of tokens corresponding to the SQL string `sql`."""
 770        self.reset()
 771        self.sql = sql
 772        self.size = len(sql)
 773        self._scan()
 774        return self.tokens
 775
 776    def _scan(self, until: t.Optional[t.Callable] = None) -> None:
 777        while self.size and not self._end:
 778            self._start = self._current
 779            self._advance()
 780
 781            if not self._char:
 782                break
 783
 784            white_space = self.WHITE_SPACE.get(self._char)  # type: ignore
 785            identifier_end = self._IDENTIFIERS.get(self._char)  # type: ignore
 786
 787            if white_space:
 788                if white_space == TokenType.BREAK:
 789                    self._col = 1
 790                    self._line += 1
 791            elif self._char.isdigit():  # type:ignore
 792                self._scan_number()
 793            elif identifier_end:
 794                self._scan_identifier(identifier_end)
 795            else:
 796                self._scan_keywords()
 797
 798            if until and until():
 799                break
 800
 801    def _chars(self, size: int) -> str:
 802        if size == 1:
 803            return self._char  # type: ignore
 804        start = self._current - 1
 805        end = start + size
 806        if end <= self.size:
 807            return self.sql[start:end]
 808        return ""
 809
 810    def _advance(self, i: int = 1) -> None:
 811        self._col += i
 812        self._current += i
 813        self._end = self._current >= self.size  # type: ignore
 814        self._char = self.sql[self._current - 1]  # type: ignore
 815        self._peek = self.sql[self._current] if self._current < self.size else ""  # type: ignore
 816
 817    @property
 818    def _text(self) -> str:
 819        return self.sql[self._start : self._current]
 820
 821    def _add(self, token_type: TokenType, text: t.Optional[str] = None) -> None:
 822        self._prev_token_line = self._line
 823        self._prev_token_comments = self._comments
 824        self._prev_token_type = token_type  # type: ignore
 825        self.tokens.append(
 826            Token(
 827                token_type,
 828                self._text if text is None else text,
 829                self._line,
 830                self._col,
 831                self._comments,
 832            )
 833        )
 834        self._comments = []
 835
 836        # If we have either a semicolon or a begin token before the command's token, we'll parse
 837        # whatever follows the command's token as a string
 838        if token_type in self.COMMANDS and (
 839            len(self.tokens) == 1 or self.tokens[-2].token_type in self.COMMAND_PREFIX_TOKENS
 840        ):
 841            start = self._current
 842            tokens = len(self.tokens)
 843            self._scan(lambda: self._peek == ";")
 844            self.tokens = self.tokens[:tokens]
 845            text = self.sql[start : self._current].strip()
 846            if text:
 847                self._add(TokenType.STRING, text)
 848
 849    def _scan_keywords(self) -> None:
 850        size = 0
 851        word = None
 852        chars: t.Optional[str] = self._text
 853        char = chars
 854        prev_space = False
 855        skip = False
 856        trie = self.KEYWORD_TRIE
 857
 858        while chars:
 859            if skip:
 860                result = 1
 861            else:
 862                result, trie = in_trie(trie, char.upper())  # type: ignore
 863
 864            if result == 0:
 865                break
 866            if result == 2:
 867                word = chars
 868            size += 1
 869            end = self._current - 1 + size
 870
 871            if end < self.size:
 872                char = self.sql[end]
 873                is_space = char in self.WHITE_SPACE
 874
 875                if not is_space or not prev_space:
 876                    if is_space:
 877                        char = " "
 878                    chars += char
 879                    prev_space = is_space
 880                    skip = False
 881                else:
 882                    skip = True
 883            else:
 884                chars = None
 885
 886        if not word:
 887            if self._char in self.SINGLE_TOKENS:
 888                self._add(self.SINGLE_TOKENS[self._char])  # type: ignore
 889                return
 890            self._scan_var()
 891            return
 892
 893        if self._scan_string(word):
 894            return
 895        if self._scan_formatted_string(word):
 896            return
 897        if self._scan_comment(word):
 898            return
 899
 900        self._advance(size - 1)
 901        self._add(self.KEYWORDS[word.upper()])
 902
 903    def _scan_comment(self, comment_start: str) -> bool:
 904        if comment_start not in self._COMMENTS:  # type: ignore
 905            return False
 906
 907        comment_start_line = self._line
 908        comment_start_size = len(comment_start)
 909        comment_end = self._COMMENTS[comment_start]  # type: ignore
 910
 911        if comment_end:
 912            comment_end_size = len(comment_end)
 913
 914            while not self._end and self._chars(comment_end_size) != comment_end:
 915                self._advance()
 916
 917            self._comments.append(self._text[comment_start_size : -comment_end_size + 1])  # type: ignore
 918            self._advance(comment_end_size - 1)
 919        else:
 920            while not self._end and self.WHITE_SPACE.get(self._peek) != TokenType.BREAK:  # type: ignore
 921                self._advance()
 922            self._comments.append(self._text[comment_start_size:])  # type: ignore
 923
 924        # Leading comment is attached to the succeeding token, whilst trailing comment to the preceding.
 925        # Multiple consecutive comments are preserved by appending them to the current comments list.
 926        if comment_start_line == self._prev_token_line:
 927            self.tokens[-1].comments.extend(self._comments)
 928            self._comments = []
 929
 930        return True
 931
 932    def _scan_number(self) -> None:
 933        if self._char == "0":
 934            peek = self._peek.upper()  # type: ignore
 935            if peek == "B":
 936                return self._scan_bits()
 937            elif peek == "X":
 938                return self._scan_hex()
 939
 940        decimal = False
 941        scientific = 0
 942
 943        while True:
 944            if self._peek.isdigit():  # type: ignore
 945                self._advance()
 946            elif self._peek == "." and not decimal:
 947                decimal = True
 948                self._advance()
 949            elif self._peek in ("-", "+") and scientific == 1:
 950                scientific += 1
 951                self._advance()
 952            elif self._peek.upper() == "E" and not scientific:  # type: ignore
 953                scientific += 1
 954                self._advance()
 955            elif self._peek.isidentifier():  # type: ignore
 956                number_text = self._text
 957                literal = []
 958
 959                while self._peek.strip() and self._peek not in self.SINGLE_TOKENS:  # type: ignore
 960                    literal.append(self._peek.upper())  # type: ignore
 961                    self._advance()
 962
 963                literal = "".join(literal)  # type: ignore
 964                token_type = self.KEYWORDS.get(self.NUMERIC_LITERALS.get(literal))  # type: ignore
 965
 966                if token_type:
 967                    self._add(TokenType.NUMBER, number_text)
 968                    self._add(TokenType.DCOLON, "::")
 969                    return self._add(token_type, literal)  # type: ignore
 970                elif self.IDENTIFIER_CAN_START_WITH_DIGIT:
 971                    return self._add(TokenType.VAR)
 972
 973                self._add(TokenType.NUMBER, number_text)
 974                return self._advance(-len(literal))
 975            else:
 976                return self._add(TokenType.NUMBER)
 977
 978    def _scan_bits(self) -> None:
 979        self._advance()
 980        value = self._extract_value()
 981        try:
 982            self._add(TokenType.BIT_STRING, f"{int(value, 2)}")
 983        except ValueError:
 984            self._add(TokenType.IDENTIFIER)
 985
 986    def _scan_hex(self) -> None:
 987        self._advance()
 988        value = self._extract_value()
 989        try:
 990            self._add(TokenType.HEX_STRING, f"{int(value, 16)}")
 991        except ValueError:
 992            self._add(TokenType.IDENTIFIER)
 993
 994    def _extract_value(self) -> str:
 995        while True:
 996            char = self._peek.strip()  # type: ignore
 997            if char and char not in self.SINGLE_TOKENS:
 998                self._advance()
 999            else:
1000                break
1001
1002        return self._text
1003
1004    def _scan_string(self, quote: str) -> bool:
1005        quote_end = self._QUOTES.get(quote)  # type: ignore
1006        if quote_end is None:
1007            return False
1008
1009        self._advance(len(quote))
1010        text = self._extract_string(quote_end)
1011        text = text.encode(self.ENCODE).decode(self.ENCODE) if self.ENCODE else text  # type: ignore
1012        self._add(TokenType.NATIONAL if quote[0].upper() == "N" else TokenType.STRING, text)
1013        return True
1014
1015    # X'1234, b'0110', E'\\\\\' etc.
1016    def _scan_formatted_string(self, string_start: str) -> bool:
1017        if string_start in self._HEX_STRINGS:  # type: ignore
1018            delimiters = self._HEX_STRINGS  # type: ignore
1019            token_type = TokenType.HEX_STRING
1020            base = 16
1021        elif string_start in self._BIT_STRINGS:  # type: ignore
1022            delimiters = self._BIT_STRINGS  # type: ignore
1023            token_type = TokenType.BIT_STRING
1024            base = 2
1025        elif string_start in self._BYTE_STRINGS:  # type: ignore
1026            delimiters = self._BYTE_STRINGS  # type: ignore
1027            token_type = TokenType.BYTE_STRING
1028            base = None
1029        else:
1030            return False
1031
1032        self._advance(len(string_start))
1033        string_end = delimiters.get(string_start)
1034        text = self._extract_string(string_end)
1035
1036        if base is None:
1037            self._add(token_type, text)
1038        else:
1039            try:
1040                self._add(token_type, f"{int(text, base)}")
1041            except:
1042                raise RuntimeError(
1043                    f"Numeric string contains invalid characters from {self._line}:{self._start}"
1044                )
1045
1046        return True
1047
1048    def _scan_identifier(self, identifier_end: str) -> None:
1049        text = ""
1050        identifier_end_is_escape = identifier_end in self._IDENTIFIER_ESCAPES
1051
1052        while True:
1053            if self._end:
1054                raise RuntimeError(f"Missing {identifier_end} from {self._line}:{self._start}")
1055
1056            self._advance()
1057            if self._char == identifier_end:
1058                if identifier_end_is_escape and self._peek == identifier_end:
1059                    text += identifier_end  # type: ignore
1060                    self._advance()
1061                    continue
1062
1063                break
1064
1065            text += self._char  # type: ignore
1066
1067        self._add(TokenType.IDENTIFIER, text)
1068
1069    def _scan_var(self) -> None:
1070        while True:
1071            char = self._peek.strip()  # type: ignore
1072            if char and char not in self.SINGLE_TOKENS:
1073                self._advance()
1074            else:
1075                break
1076        self._add(
1077            TokenType.VAR
1078            if self._prev_token_type == TokenType.PARAMETER
1079            else self.KEYWORDS.get(self._text.upper(), TokenType.VAR)
1080        )
1081
1082    def _extract_string(self, delimiter: str) -> str:
1083        text = ""
1084        delim_size = len(delimiter)
1085
1086        while True:
1087            if self._char in self._STRING_ESCAPES and (
1088                self._peek == delimiter or self._peek in self._STRING_ESCAPES
1089            ):
1090                if self._peek == delimiter:
1091                    text += self._peek  # type: ignore
1092                else:
1093                    text += self._char + self._peek  # type: ignore
1094
1095                if self._current + 1 < self.size:
1096                    self._advance(2)
1097                else:
1098                    raise RuntimeError(f"Missing {delimiter} from {self._line}:{self._current}")
1099            else:
1100                if self._chars(delim_size) == delimiter:
1101                    if delim_size > 1:
1102                        self._advance(delim_size - 1)
1103                    break
1104
1105                if self._end:
1106                    raise RuntimeError(f"Missing {delimiter} from {self._line}:{self._start}")
1107                text += self._char  # type: ignore
1108                self._advance()
1109
1110        return text