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