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    LT_AT = auto()
  50    AT_GT = auto()
  51    DOLLAR = auto()
  52    PARAMETER = auto()
  53    SESSION_PARAMETER = auto()
  54    NATIONAL = auto()
  55    DAMP = auto()
  56
  57    BLOCK_START = auto()
  58    BLOCK_END = auto()
  59
  60    SPACE = auto()
  61    BREAK = auto()
  62
  63    STRING = auto()
  64    NUMBER = auto()
  65    IDENTIFIER = auto()
  66    DATABASE = auto()
  67    COLUMN = auto()
  68    COLUMN_DEF = auto()
  69    SCHEMA = auto()
  70    TABLE = auto()
  71    VAR = auto()
  72    BIT_STRING = auto()
  73    HEX_STRING = auto()
  74    BYTE_STRING = auto()
  75
  76    # types
  77    BIT = auto()
  78    BOOLEAN = auto()
  79    TINYINT = auto()
  80    UTINYINT = auto()
  81    SMALLINT = auto()
  82    USMALLINT = auto()
  83    INT = auto()
  84    UINT = auto()
  85    BIGINT = auto()
  86    UBIGINT = auto()
  87    FLOAT = auto()
  88    DOUBLE = auto()
  89    DECIMAL = auto()
  90    CHAR = auto()
  91    NCHAR = auto()
  92    VARCHAR = auto()
  93    NVARCHAR = auto()
  94    TEXT = auto()
  95    MEDIUMTEXT = auto()
  96    LONGTEXT = auto()
  97    MEDIUMBLOB = auto()
  98    LONGBLOB = auto()
  99    BINARY = auto()
 100    VARBINARY = auto()
 101    JSON = auto()
 102    JSONB = auto()
 103    TIME = auto()
 104    TIMESTAMP = auto()
 105    TIMESTAMPTZ = auto()
 106    TIMESTAMPLTZ = auto()
 107    DATETIME = auto()
 108    DATE = auto()
 109    UUID = auto()
 110    GEOGRAPHY = auto()
 111    NULLABLE = auto()
 112    GEOMETRY = auto()
 113    HLLSKETCH = auto()
 114    HSTORE = auto()
 115    SUPER = auto()
 116    SERIAL = auto()
 117    SMALLSERIAL = auto()
 118    BIGSERIAL = auto()
 119    XML = auto()
 120    UNIQUEIDENTIFIER = auto()
 121    MONEY = auto()
 122    SMALLMONEY = auto()
 123    ROWVERSION = auto()
 124    IMAGE = auto()
 125    VARIANT = auto()
 126    OBJECT = auto()
 127    INET = auto()
 128
 129    # keywords
 130    ALIAS = auto()
 131    ALTER = auto()
 132    ALWAYS = auto()
 133    ALL = auto()
 134    ANTI = auto()
 135    ANY = auto()
 136    APPLY = auto()
 137    ARRAY = auto()
 138    ASC = auto()
 139    ASOF = auto()
 140    AT_TIME_ZONE = auto()
 141    AUTO_INCREMENT = auto()
 142    BEGIN = auto()
 143    BETWEEN = auto()
 144    BOTH = auto()
 145    BUCKET = auto()
 146    BY_DEFAULT = auto()
 147    CACHE = auto()
 148    CASCADE = auto()
 149    CASE = auto()
 150    CHARACTER_SET = auto()
 151    CLUSTER_BY = auto()
 152    COLLATE = auto()
 153    COMMAND = auto()
 154    COMMENT = auto()
 155    COMMIT = auto()
 156    COMPOUND = auto()
 157    CONSTRAINT = auto()
 158    CREATE = auto()
 159    CROSS = auto()
 160    CUBE = auto()
 161    CURRENT_DATE = auto()
 162    CURRENT_DATETIME = auto()
 163    CURRENT_ROW = auto()
 164    CURRENT_TIME = auto()
 165    CURRENT_TIMESTAMP = auto()
 166    DEFAULT = auto()
 167    DELETE = auto()
 168    DESC = auto()
 169    DESCRIBE = auto()
 170    DISTINCT = auto()
 171    DISTINCT_FROM = auto()
 172    DISTRIBUTE_BY = auto()
 173    DIV = auto()
 174    DROP = auto()
 175    ELSE = auto()
 176    END = auto()
 177    ESCAPE = auto()
 178    EXCEPT = auto()
 179    EXECUTE = auto()
 180    EXISTS = auto()
 181    FALSE = auto()
 182    FETCH = auto()
 183    FILTER = auto()
 184    FINAL = auto()
 185    FIRST = auto()
 186    FOLLOWING = auto()
 187    FOR = auto()
 188    FOREIGN_KEY = auto()
 189    FORMAT = auto()
 190    FROM = auto()
 191    FULL = auto()
 192    FUNCTION = auto()
 193    GLOB = auto()
 194    GLOBAL = auto()
 195    GROUP_BY = auto()
 196    GROUPING_SETS = auto()
 197    HAVING = auto()
 198    HINT = auto()
 199    IF = auto()
 200    IGNORE_NULLS = auto()
 201    ILIKE = auto()
 202    ILIKE_ANY = auto()
 203    IN = auto()
 204    INDEX = auto()
 205    INNER = auto()
 206    INSERT = auto()
 207    INTERSECT = auto()
 208    INTERVAL = auto()
 209    INTO = auto()
 210    INTRODUCER = auto()
 211    IRLIKE = auto()
 212    IS = auto()
 213    ISNULL = auto()
 214    JOIN = auto()
 215    JOIN_MARKER = auto()
 216    LANGUAGE = auto()
 217    LATERAL = auto()
 218    LAZY = auto()
 219    LEADING = auto()
 220    LEFT = auto()
 221    LIKE = auto()
 222    LIKE_ANY = auto()
 223    LIMIT = auto()
 224    LOAD_DATA = auto()
 225    LOCAL = auto()
 226    MAP = auto()
 227    MATCH_RECOGNIZE = auto()
 228    MATERIALIZED = auto()
 229    MERGE = auto()
 230    MOD = auto()
 231    NATURAL = auto()
 232    NEXT = auto()
 233    NO_ACTION = auto()
 234    NOTNULL = auto()
 235    NULL = auto()
 236    NULLS_FIRST = auto()
 237    NULLS_LAST = auto()
 238    OFFSET = auto()
 239    ON = auto()
 240    ONLY = auto()
 241    OPTIONS = auto()
 242    ORDER_BY = auto()
 243    ORDERED = auto()
 244    ORDINALITY = auto()
 245    OUTER = auto()
 246    OUT_OF = auto()
 247    OVER = auto()
 248    OVERLAPS = auto()
 249    OVERWRITE = auto()
 250    PARTITION = auto()
 251    PARTITION_BY = auto()
 252    PERCENT = auto()
 253    PIVOT = auto()
 254    PLACEHOLDER = auto()
 255    PRECEDING = auto()
 256    PRIMARY_KEY = auto()
 257    PROCEDURE = auto()
 258    PROPERTIES = auto()
 259    PSEUDO_TYPE = auto()
 260    QUALIFY = auto()
 261    QUOTE = auto()
 262    RANGE = auto()
 263    RECURSIVE = auto()
 264    REPLACE = auto()
 265    RESPECT_NULLS = auto()
 266    RETURNING = auto()
 267    REFERENCES = auto()
 268    RIGHT = auto()
 269    RLIKE = auto()
 270    ROLLBACK = auto()
 271    ROLLUP = auto()
 272    ROW = auto()
 273    ROWS = auto()
 274    SEED = auto()
 275    SELECT = auto()
 276    SEMI = auto()
 277    SEPARATOR = auto()
 278    SERDE_PROPERTIES = auto()
 279    SET = auto()
 280    SHOW = auto()
 281    SIMILAR_TO = auto()
 282    SOME = auto()
 283    SORTKEY = auto()
 284    SORT_BY = auto()
 285    STRUCT = auto()
 286    TABLE_SAMPLE = auto()
 287    TEMPORARY = auto()
 288    TOP = auto()
 289    THEN = auto()
 290    TRAILING = auto()
 291    TRUE = auto()
 292    UNBOUNDED = auto()
 293    UNCACHE = auto()
 294    UNION = auto()
 295    UNLOGGED = auto()
 296    UNNEST = auto()
 297    UNPIVOT = auto()
 298    UPDATE = auto()
 299    USE = auto()
 300    USING = auto()
 301    VALUES = auto()
 302    VIEW = auto()
 303    VOLATILE = auto()
 304    WHEN = auto()
 305    WHERE = auto()
 306    WINDOW = auto()
 307    WITH = auto()
 308    WITH_TIME_ZONE = auto()
 309    WITH_LOCAL_TIME_ZONE = auto()
 310    WITHIN_GROUP = auto()
 311    WITHOUT_TIME_ZONE = auto()
 312    UNIQUE = auto()
 313
 314
 315class Token:
 316    __slots__ = ("token_type", "text", "line", "col", "comments")
 317
 318    @classmethod
 319    def number(cls, number: int) -> Token:
 320        """Returns a NUMBER token with `number` as its text."""
 321        return cls(TokenType.NUMBER, str(number))
 322
 323    @classmethod
 324    def string(cls, string: str) -> Token:
 325        """Returns a STRING token with `string` as its text."""
 326        return cls(TokenType.STRING, string)
 327
 328    @classmethod
 329    def identifier(cls, identifier: str) -> Token:
 330        """Returns an IDENTIFIER token with `identifier` as its text."""
 331        return cls(TokenType.IDENTIFIER, identifier)
 332
 333    @classmethod
 334    def var(cls, var: str) -> Token:
 335        """Returns an VAR token with `var` as its text."""
 336        return cls(TokenType.VAR, var)
 337
 338    def __init__(
 339        self,
 340        token_type: TokenType,
 341        text: str,
 342        line: int = 1,
 343        col: int = 1,
 344        comments: t.List[str] = [],
 345    ) -> None:
 346        self.token_type = token_type
 347        self.text = text
 348        self.line = line
 349        self.col = max(col - len(text), 1)
 350        self.comments = comments
 351
 352    def __repr__(self) -> str:
 353        attributes = ", ".join(f"{k}: {getattr(self, k)}" for k in self.__slots__)
 354        return f"<Token {attributes}>"
 355
 356
 357class _Tokenizer(type):
 358    def __new__(cls, clsname, bases, attrs):  # type: ignore
 359        klass = super().__new__(cls, clsname, bases, attrs)
 360
 361        klass._QUOTES = {
 362            f"{prefix}{s}": e
 363            for s, e in cls._delimeter_list_to_dict(klass.QUOTES).items()
 364            for prefix in (("",) if s[0].isalpha() else ("", "n", "N"))
 365        }
 366        klass._BIT_STRINGS = cls._delimeter_list_to_dict(klass.BIT_STRINGS)
 367        klass._HEX_STRINGS = cls._delimeter_list_to_dict(klass.HEX_STRINGS)
 368        klass._BYTE_STRINGS = cls._delimeter_list_to_dict(klass.BYTE_STRINGS)
 369        klass._IDENTIFIERS = cls._delimeter_list_to_dict(klass.IDENTIFIERS)
 370        klass._STRING_ESCAPES = set(klass.STRING_ESCAPES)
 371        klass._IDENTIFIER_ESCAPES = set(klass.IDENTIFIER_ESCAPES)
 372        klass._COMMENTS = dict(
 373            (comment, None) if isinstance(comment, str) else (comment[0], comment[1])
 374            for comment in klass.COMMENTS
 375        )
 376
 377        klass.KEYWORD_TRIE = new_trie(
 378            key.upper()
 379            for key in {
 380                **klass.KEYWORDS,
 381                **{comment: TokenType.COMMENT for comment in klass._COMMENTS},
 382                **{quote: TokenType.QUOTE for quote in klass._QUOTES},
 383                **{bit_string: TokenType.BIT_STRING for bit_string in klass._BIT_STRINGS},
 384                **{hex_string: TokenType.HEX_STRING for hex_string in klass._HEX_STRINGS},
 385                **{byte_string: TokenType.BYTE_STRING for byte_string in klass._BYTE_STRINGS},
 386            }
 387            if " " in key or any(single in key for single in klass.SINGLE_TOKENS)
 388        )
 389
 390        return klass
 391
 392    @staticmethod
 393    def _delimeter_list_to_dict(list: t.List[str | t.Tuple[str, str]]) -> t.Dict[str, str]:
 394        return dict((item, item) if isinstance(item, str) else (item[0], item[1]) for item in list)
 395
 396
 397class Tokenizer(metaclass=_Tokenizer):
 398    SINGLE_TOKENS = {
 399        "(": TokenType.L_PAREN,
 400        ")": TokenType.R_PAREN,
 401        "[": TokenType.L_BRACKET,
 402        "]": TokenType.R_BRACKET,
 403        "{": TokenType.L_BRACE,
 404        "}": TokenType.R_BRACE,
 405        "&": TokenType.AMP,
 406        "^": TokenType.CARET,
 407        ":": TokenType.COLON,
 408        ",": TokenType.COMMA,
 409        ".": TokenType.DOT,
 410        "-": TokenType.DASH,
 411        "=": TokenType.EQ,
 412        ">": TokenType.GT,
 413        "<": TokenType.LT,
 414        "%": TokenType.MOD,
 415        "!": TokenType.NOT,
 416        "|": TokenType.PIPE,
 417        "+": TokenType.PLUS,
 418        ";": TokenType.SEMICOLON,
 419        "/": TokenType.SLASH,
 420        "\\": TokenType.BACKSLASH,
 421        "*": TokenType.STAR,
 422        "~": TokenType.TILDA,
 423        "?": TokenType.PLACEHOLDER,
 424        "@": TokenType.PARAMETER,
 425        # used for breaking a var like x'y' but nothing else
 426        # the token type doesn't matter
 427        "'": TokenType.QUOTE,
 428        "`": TokenType.IDENTIFIER,
 429        '"': TokenType.IDENTIFIER,
 430        "#": TokenType.HASH,
 431    }
 432
 433    QUOTES: t.List[t.Tuple[str, str] | str] = ["'"]
 434
 435    BIT_STRINGS: t.List[str | t.Tuple[str, str]] = []
 436
 437    HEX_STRINGS: t.List[str | t.Tuple[str, str]] = []
 438
 439    BYTE_STRINGS: t.List[str | t.Tuple[str, str]] = []
 440
 441    IDENTIFIERS: t.List[str | t.Tuple[str, str]] = ['"']
 442
 443    STRING_ESCAPES = ["'"]
 444
 445    _STRING_ESCAPES: t.Set[str] = set()
 446
 447    IDENTIFIER_ESCAPES = ['"']
 448
 449    _IDENTIFIER_ESCAPES: t.Set[str] = set()
 450
 451    KEYWORDS = {
 452        **{f"{{%{postfix}": TokenType.BLOCK_START for postfix in ("", "+", "-")},
 453        **{f"{prefix}%}}": TokenType.BLOCK_END for prefix in ("", "+", "-")},
 454        "{{+": TokenType.BLOCK_START,
 455        "{{-": TokenType.BLOCK_START,
 456        "+}}": TokenType.BLOCK_END,
 457        "-}}": TokenType.BLOCK_END,
 458        "/*+": TokenType.HINT,
 459        "==": TokenType.EQ,
 460        "::": TokenType.DCOLON,
 461        "||": TokenType.DPIPE,
 462        ">=": TokenType.GTE,
 463        "<=": TokenType.LTE,
 464        "<>": TokenType.NEQ,
 465        "!=": TokenType.NEQ,
 466        "<=>": TokenType.NULLSAFE_EQ,
 467        "->": TokenType.ARROW,
 468        "->>": TokenType.DARROW,
 469        "=>": TokenType.FARROW,
 470        "#>": TokenType.HASH_ARROW,
 471        "#>>": TokenType.DHASH_ARROW,
 472        "<->": TokenType.LR_ARROW,
 473        "&&": TokenType.DAMP,
 474        "ALL": TokenType.ALL,
 475        "ALWAYS": TokenType.ALWAYS,
 476        "AND": TokenType.AND,
 477        "ANTI": TokenType.ANTI,
 478        "ANY": TokenType.ANY,
 479        "ASC": TokenType.ASC,
 480        "AS": TokenType.ALIAS,
 481        "AT TIME ZONE": TokenType.AT_TIME_ZONE,
 482        "AUTOINCREMENT": TokenType.AUTO_INCREMENT,
 483        "AUTO_INCREMENT": TokenType.AUTO_INCREMENT,
 484        "BEGIN": TokenType.BEGIN,
 485        "BETWEEN": TokenType.BETWEEN,
 486        "BOTH": TokenType.BOTH,
 487        "BUCKET": TokenType.BUCKET,
 488        "BY DEFAULT": TokenType.BY_DEFAULT,
 489        "CACHE": TokenType.CACHE,
 490        "UNCACHE": TokenType.UNCACHE,
 491        "CASE": TokenType.CASE,
 492        "CASCADE": TokenType.CASCADE,
 493        "CHARACTER SET": TokenType.CHARACTER_SET,
 494        "CLUSTER BY": TokenType.CLUSTER_BY,
 495        "COLLATE": TokenType.COLLATE,
 496        "COLUMN": TokenType.COLUMN,
 497        "COMMIT": TokenType.COMMIT,
 498        "COMPOUND": TokenType.COMPOUND,
 499        "CONSTRAINT": TokenType.CONSTRAINT,
 500        "CREATE": TokenType.CREATE,
 501        "CROSS": TokenType.CROSS,
 502        "CUBE": TokenType.CUBE,
 503        "CURRENT_DATE": TokenType.CURRENT_DATE,
 504        "CURRENT ROW": TokenType.CURRENT_ROW,
 505        "CURRENT_TIME": TokenType.CURRENT_TIME,
 506        "CURRENT_TIMESTAMP": TokenType.CURRENT_TIMESTAMP,
 507        "DATABASE": TokenType.DATABASE,
 508        "DEFAULT": TokenType.DEFAULT,
 509        "DELETE": TokenType.DELETE,
 510        "DESC": TokenType.DESC,
 511        "DESCRIBE": TokenType.DESCRIBE,
 512        "DISTINCT": TokenType.DISTINCT,
 513        "DISTINCT FROM": TokenType.DISTINCT_FROM,
 514        "DISTRIBUTE BY": TokenType.DISTRIBUTE_BY,
 515        "DIV": TokenType.DIV,
 516        "DROP": TokenType.DROP,
 517        "ELSE": TokenType.ELSE,
 518        "END": TokenType.END,
 519        "ESCAPE": TokenType.ESCAPE,
 520        "EXCEPT": TokenType.EXCEPT,
 521        "EXECUTE": TokenType.EXECUTE,
 522        "EXISTS": TokenType.EXISTS,
 523        "FALSE": TokenType.FALSE,
 524        "FETCH": TokenType.FETCH,
 525        "FILTER": TokenType.FILTER,
 526        "FIRST": TokenType.FIRST,
 527        "FULL": TokenType.FULL,
 528        "FUNCTION": TokenType.FUNCTION,
 529        "FOLLOWING": TokenType.FOLLOWING,
 530        "FOR": TokenType.FOR,
 531        "FOREIGN KEY": TokenType.FOREIGN_KEY,
 532        "FORMAT": TokenType.FORMAT,
 533        "FROM": TokenType.FROM,
 534        "GLOB": TokenType.GLOB,
 535        "GROUP BY": TokenType.GROUP_BY,
 536        "GROUPING SETS": TokenType.GROUPING_SETS,
 537        "HAVING": TokenType.HAVING,
 538        "IF": TokenType.IF,
 539        "ILIKE": TokenType.ILIKE,
 540        "IGNORE NULLS": TokenType.IGNORE_NULLS,
 541        "IN": TokenType.IN,
 542        "INDEX": TokenType.INDEX,
 543        "INET": TokenType.INET,
 544        "INNER": TokenType.INNER,
 545        "INSERT": TokenType.INSERT,
 546        "INTERVAL": TokenType.INTERVAL,
 547        "INTERSECT": TokenType.INTERSECT,
 548        "INTO": TokenType.INTO,
 549        "IS": TokenType.IS,
 550        "ISNULL": TokenType.ISNULL,
 551        "JOIN": TokenType.JOIN,
 552        "LATERAL": TokenType.LATERAL,
 553        "LAZY": TokenType.LAZY,
 554        "LEADING": TokenType.LEADING,
 555        "LEFT": TokenType.LEFT,
 556        "LIKE": TokenType.LIKE,
 557        "LIMIT": TokenType.LIMIT,
 558        "LOAD DATA": TokenType.LOAD_DATA,
 559        "LOCAL": TokenType.LOCAL,
 560        "MATERIALIZED": TokenType.MATERIALIZED,
 561        "MERGE": TokenType.MERGE,
 562        "NATURAL": TokenType.NATURAL,
 563        "NEXT": TokenType.NEXT,
 564        "NO ACTION": TokenType.NO_ACTION,
 565        "NOT": TokenType.NOT,
 566        "NOTNULL": TokenType.NOTNULL,
 567        "NULL": TokenType.NULL,
 568        "NULLS FIRST": TokenType.NULLS_FIRST,
 569        "NULLS LAST": TokenType.NULLS_LAST,
 570        "OBJECT": TokenType.OBJECT,
 571        "OFFSET": TokenType.OFFSET,
 572        "ON": TokenType.ON,
 573        "ONLY": TokenType.ONLY,
 574        "OPTIONS": TokenType.OPTIONS,
 575        "OR": TokenType.OR,
 576        "ORDER BY": TokenType.ORDER_BY,
 577        "ORDINALITY": TokenType.ORDINALITY,
 578        "OUTER": TokenType.OUTER,
 579        "OUT OF": TokenType.OUT_OF,
 580        "OVER": TokenType.OVER,
 581        "OVERLAPS": TokenType.OVERLAPS,
 582        "OVERWRITE": TokenType.OVERWRITE,
 583        "PARTITION": TokenType.PARTITION,
 584        "PARTITION BY": TokenType.PARTITION_BY,
 585        "PARTITIONED BY": TokenType.PARTITION_BY,
 586        "PARTITIONED_BY": TokenType.PARTITION_BY,
 587        "PERCENT": TokenType.PERCENT,
 588        "PIVOT": TokenType.PIVOT,
 589        "PRECEDING": TokenType.PRECEDING,
 590        "PRIMARY KEY": TokenType.PRIMARY_KEY,
 591        "PROCEDURE": TokenType.PROCEDURE,
 592        "QUALIFY": TokenType.QUALIFY,
 593        "RANGE": TokenType.RANGE,
 594        "RECURSIVE": TokenType.RECURSIVE,
 595        "REGEXP": TokenType.RLIKE,
 596        "REPLACE": TokenType.REPLACE,
 597        "RESPECT NULLS": TokenType.RESPECT_NULLS,
 598        "REFERENCES": TokenType.REFERENCES,
 599        "RIGHT": TokenType.RIGHT,
 600        "RLIKE": TokenType.RLIKE,
 601        "ROLLBACK": TokenType.ROLLBACK,
 602        "ROLLUP": TokenType.ROLLUP,
 603        "ROW": TokenType.ROW,
 604        "ROWS": TokenType.ROWS,
 605        "SCHEMA": TokenType.SCHEMA,
 606        "SEED": TokenType.SEED,
 607        "SELECT": TokenType.SELECT,
 608        "SEMI": TokenType.SEMI,
 609        "SET": TokenType.SET,
 610        "SHOW": TokenType.SHOW,
 611        "SIMILAR TO": TokenType.SIMILAR_TO,
 612        "SOME": TokenType.SOME,
 613        "SORTKEY": TokenType.SORTKEY,
 614        "SORT BY": TokenType.SORT_BY,
 615        "TABLE": TokenType.TABLE,
 616        "TABLESAMPLE": TokenType.TABLE_SAMPLE,
 617        "TEMP": TokenType.TEMPORARY,
 618        "TEMPORARY": TokenType.TEMPORARY,
 619        "THEN": TokenType.THEN,
 620        "TRUE": TokenType.TRUE,
 621        "TRAILING": TokenType.TRAILING,
 622        "UNBOUNDED": TokenType.UNBOUNDED,
 623        "UNION": TokenType.UNION,
 624        "UNLOGGED": TokenType.UNLOGGED,
 625        "UNNEST": TokenType.UNNEST,
 626        "UNPIVOT": TokenType.UNPIVOT,
 627        "UPDATE": TokenType.UPDATE,
 628        "USE": TokenType.USE,
 629        "USING": TokenType.USING,
 630        "VALUES": TokenType.VALUES,
 631        "VIEW": TokenType.VIEW,
 632        "VOLATILE": TokenType.VOLATILE,
 633        "WHEN": TokenType.WHEN,
 634        "WHERE": TokenType.WHERE,
 635        "WINDOW": TokenType.WINDOW,
 636        "WITH": TokenType.WITH,
 637        "WITH TIME ZONE": TokenType.WITH_TIME_ZONE,
 638        "WITH LOCAL TIME ZONE": TokenType.WITH_LOCAL_TIME_ZONE,
 639        "WITHIN GROUP": TokenType.WITHIN_GROUP,
 640        "WITHOUT TIME ZONE": TokenType.WITHOUT_TIME_ZONE,
 641        "APPLY": TokenType.APPLY,
 642        "ARRAY": TokenType.ARRAY,
 643        "BIT": TokenType.BIT,
 644        "BOOL": TokenType.BOOLEAN,
 645        "BOOLEAN": TokenType.BOOLEAN,
 646        "BYTE": TokenType.TINYINT,
 647        "TINYINT": TokenType.TINYINT,
 648        "SHORT": TokenType.SMALLINT,
 649        "SMALLINT": TokenType.SMALLINT,
 650        "INT2": TokenType.SMALLINT,
 651        "INTEGER": TokenType.INT,
 652        "INT": TokenType.INT,
 653        "INT4": TokenType.INT,
 654        "LONG": TokenType.BIGINT,
 655        "BIGINT": TokenType.BIGINT,
 656        "INT8": TokenType.BIGINT,
 657        "DECIMAL": TokenType.DECIMAL,
 658        "MAP": TokenType.MAP,
 659        "NULLABLE": TokenType.NULLABLE,
 660        "NUMBER": TokenType.DECIMAL,
 661        "NUMERIC": TokenType.DECIMAL,
 662        "FIXED": TokenType.DECIMAL,
 663        "REAL": TokenType.FLOAT,
 664        "FLOAT": TokenType.FLOAT,
 665        "FLOAT4": TokenType.FLOAT,
 666        "FLOAT8": TokenType.DOUBLE,
 667        "DOUBLE": TokenType.DOUBLE,
 668        "DOUBLE PRECISION": TokenType.DOUBLE,
 669        "JSON": TokenType.JSON,
 670        "CHAR": TokenType.CHAR,
 671        "CHARACTER": TokenType.CHAR,
 672        "NCHAR": TokenType.NCHAR,
 673        "VARCHAR": TokenType.VARCHAR,
 674        "VARCHAR2": TokenType.VARCHAR,
 675        "NVARCHAR": TokenType.NVARCHAR,
 676        "NVARCHAR2": TokenType.NVARCHAR,
 677        "STR": TokenType.TEXT,
 678        "STRING": TokenType.TEXT,
 679        "TEXT": TokenType.TEXT,
 680        "CLOB": TokenType.TEXT,
 681        "LONGVARCHAR": TokenType.TEXT,
 682        "BINARY": TokenType.BINARY,
 683        "BLOB": TokenType.VARBINARY,
 684        "BYTEA": TokenType.VARBINARY,
 685        "VARBINARY": TokenType.VARBINARY,
 686        "TIME": TokenType.TIME,
 687        "TIMESTAMP": TokenType.TIMESTAMP,
 688        "TIMESTAMPTZ": TokenType.TIMESTAMPTZ,
 689        "TIMESTAMPLTZ": TokenType.TIMESTAMPLTZ,
 690        "DATE": TokenType.DATE,
 691        "DATETIME": TokenType.DATETIME,
 692        "UNIQUE": TokenType.UNIQUE,
 693        "STRUCT": TokenType.STRUCT,
 694        "VARIANT": TokenType.VARIANT,
 695        "ALTER": TokenType.ALTER,
 696        "ALTER AGGREGATE": TokenType.COMMAND,
 697        "ALTER DEFAULT": TokenType.COMMAND,
 698        "ALTER DOMAIN": TokenType.COMMAND,
 699        "ALTER ROLE": TokenType.COMMAND,
 700        "ALTER RULE": TokenType.COMMAND,
 701        "ALTER SEQUENCE": TokenType.COMMAND,
 702        "ALTER TYPE": TokenType.COMMAND,
 703        "ALTER USER": TokenType.COMMAND,
 704        "ALTER VIEW": TokenType.COMMAND,
 705        "ANALYZE": TokenType.COMMAND,
 706        "CALL": TokenType.COMMAND,
 707        "COMMENT": TokenType.COMMENT,
 708        "COPY": TokenType.COMMAND,
 709        "EXPLAIN": TokenType.COMMAND,
 710        "GRANT": TokenType.COMMAND,
 711        "OPTIMIZE": TokenType.COMMAND,
 712        "PREPARE": TokenType.COMMAND,
 713        "TRUNCATE": TokenType.COMMAND,
 714        "VACUUM": TokenType.COMMAND,
 715    }
 716
 717    WHITE_SPACE: t.Dict[str, TokenType] = {
 718        " ": TokenType.SPACE,
 719        "\t": TokenType.SPACE,
 720        "\n": TokenType.BREAK,
 721        "\r": TokenType.BREAK,
 722        "\r\n": TokenType.BREAK,
 723    }
 724
 725    COMMANDS = {
 726        TokenType.COMMAND,
 727        TokenType.EXECUTE,
 728        TokenType.FETCH,
 729        TokenType.SHOW,
 730    }
 731
 732    COMMAND_PREFIX_TOKENS = {TokenType.SEMICOLON, TokenType.BEGIN}
 733
 734    # handle numeric literals like in hive (3L = BIGINT)
 735    NUMERIC_LITERALS: t.Dict[str, str] = {}
 736    ENCODE: t.Optional[str] = None
 737
 738    COMMENTS = ["--", ("/*", "*/"), ("{#", "#}")]
 739    KEYWORD_TRIE = None  # autofilled
 740
 741    IDENTIFIER_CAN_START_WITH_DIGIT = False
 742
 743    __slots__ = (
 744        "sql",
 745        "size",
 746        "tokens",
 747        "_start",
 748        "_current",
 749        "_line",
 750        "_col",
 751        "_comments",
 752        "_char",
 753        "_end",
 754        "_peek",
 755        "_prev_token_line",
 756        "_prev_token_comments",
 757        "_prev_token_type",
 758    )
 759
 760    def __init__(self) -> None:
 761        self.reset()
 762
 763    def reset(self) -> None:
 764        self.sql = ""
 765        self.size = 0
 766        self.tokens: t.List[Token] = []
 767        self._start = 0
 768        self._current = 0
 769        self._line = 1
 770        self._col = 1
 771        self._comments: t.List[str] = []
 772
 773        self._char = None
 774        self._end = None
 775        self._peek = None
 776        self._prev_token_line = -1
 777        self._prev_token_comments: t.List[str] = []
 778        self._prev_token_type = None
 779
 780    def tokenize(self, sql: str) -> t.List[Token]:
 781        """Returns a list of tokens corresponding to the SQL string `sql`."""
 782        self.reset()
 783        self.sql = sql
 784        self.size = len(sql)
 785        self._scan()
 786        return self.tokens
 787
 788    def _scan(self, until: t.Optional[t.Callable] = None) -> None:
 789        while self.size and not self._end:
 790            self._start = self._current
 791            self._advance()
 792
 793            if self._char is None:
 794                break
 795
 796            if self._char not in self.WHITE_SPACE:
 797                if self._char.isdigit():
 798                    self._scan_number()
 799                elif self._char in self._IDENTIFIERS:
 800                    self._scan_identifier(self._IDENTIFIERS[self._char])
 801                else:
 802                    self._scan_keywords()
 803
 804            if until and until():
 805                break
 806
 807    def _chars(self, size: int) -> str:
 808        if size == 1:
 809            return self._char  # type: ignore
 810        start = self._current - 1
 811        end = start + size
 812        if end <= self.size:
 813            return self.sql[start:end]
 814        return ""
 815
 816    def _line_break(self, char: t.Optional[str]) -> bool:
 817        return self.WHITE_SPACE.get(char) == TokenType.BREAK  # type: ignore
 818
 819    def _advance(self, i: int = 1) -> None:
 820        if self._line_break(self._char):
 821            self._set_new_line()
 822
 823        self._col += i
 824        self._current += i
 825        self._end = self._current >= self.size  # type: ignore
 826        self._char = self.sql[self._current - 1]  # type: ignore
 827        self._peek = self.sql[self._current] if self._current < self.size else ""  # type: ignore
 828
 829    def _set_new_line(self) -> None:
 830        self._col = 1
 831        self._line += 1
 832
 833    @property
 834    def _text(self) -> str:
 835        return self.sql[self._start : self._current]
 836
 837    def _add(self, token_type: TokenType, text: t.Optional[str] = None) -> None:
 838        self._prev_token_line = self._line
 839        self._prev_token_comments = self._comments
 840        self._prev_token_type = token_type  # type: ignore
 841        self.tokens.append(
 842            Token(
 843                token_type,
 844                self._text if text is None else text,
 845                self._line,
 846                self._col,
 847                self._comments,
 848            )
 849        )
 850        self._comments = []
 851
 852        # If we have either a semicolon or a begin token before the command's token, we'll parse
 853        # whatever follows the command's token as a string
 854        if (
 855            token_type in self.COMMANDS
 856            and self._peek != ";"
 857            and (len(self.tokens) == 1 or self.tokens[-2].token_type in self.COMMAND_PREFIX_TOKENS)
 858        ):
 859            start = self._current
 860            tokens = len(self.tokens)
 861            self._scan(lambda: self._peek == ";")
 862            self.tokens = self.tokens[:tokens]
 863            text = self.sql[start : self._current].strip()
 864            if text:
 865                self._add(TokenType.STRING, text)
 866
 867    def _scan_keywords(self) -> None:
 868        size = 0
 869        word = None
 870        chars = self._text
 871        char = chars
 872        prev_space = False
 873        skip = False
 874        trie = self.KEYWORD_TRIE
 875        single_token = char in self.SINGLE_TOKENS
 876
 877        while chars:
 878            if skip:
 879                result = 1
 880            else:
 881                result, trie = in_trie(trie, char.upper())  # type: ignore
 882
 883            if result == 0:
 884                break
 885            if result == 2:
 886                word = chars
 887            size += 1
 888            end = self._current - 1 + size
 889
 890            if end < self.size:
 891                char = self.sql[end]
 892                single_token = single_token or char in self.SINGLE_TOKENS
 893                is_space = char in self.WHITE_SPACE
 894
 895                if not is_space or not prev_space:
 896                    if is_space:
 897                        char = " "
 898                    chars += char
 899                    prev_space = is_space
 900                    skip = False
 901                else:
 902                    skip = True
 903            else:
 904                chars = " "
 905
 906        word = None if not single_token and chars[-1] not in self.WHITE_SPACE else word
 907
 908        if not word:
 909            if self._char in self.SINGLE_TOKENS:
 910                self._add(self.SINGLE_TOKENS[self._char])  # type: ignore
 911                return
 912            self._scan_var()
 913            return
 914
 915        if self._scan_string(word):
 916            return
 917        if self._scan_formatted_string(word):
 918            return
 919        if self._scan_comment(word):
 920            return
 921
 922        self._advance(size - 1)
 923        self._add(self.KEYWORDS[word.upper()])
 924
 925    def _scan_comment(self, comment_start: str) -> bool:
 926        if comment_start not in self._COMMENTS:  # type: ignore
 927            return False
 928
 929        comment_start_line = self._line
 930        comment_start_size = len(comment_start)
 931        comment_end = self._COMMENTS[comment_start]  # type: ignore
 932
 933        if comment_end:
 934            comment_end_size = len(comment_end)
 935
 936            while not self._end and self._chars(comment_end_size) != comment_end:
 937                self._advance()
 938
 939            self._comments.append(self._text[comment_start_size : -comment_end_size + 1])  # type: ignore
 940            self._advance(comment_end_size - 1)
 941        else:
 942            while not self._end and not self._line_break(self._peek):
 943                self._advance()
 944            self._comments.append(self._text[comment_start_size:])  # type: ignore
 945
 946        # Leading comment is attached to the succeeding token, whilst trailing comment to the preceding.
 947        # Multiple consecutive comments are preserved by appending them to the current comments list.
 948        if comment_start_line == self._prev_token_line:
 949            self.tokens[-1].comments.extend(self._comments)
 950            self._comments = []
 951            self._prev_token_line = self._line
 952
 953        return True
 954
 955    def _scan_number(self) -> None:
 956        if self._char == "0":
 957            peek = self._peek.upper()  # type: ignore
 958            if peek == "B":
 959                return self._scan_bits()
 960            elif peek == "X":
 961                return self._scan_hex()
 962
 963        decimal = False
 964        scientific = 0
 965
 966        while True:
 967            if self._peek.isdigit():  # type: ignore
 968                self._advance()
 969            elif self._peek == "." and not decimal:
 970                decimal = True
 971                self._advance()
 972            elif self._peek in ("-", "+") and scientific == 1:
 973                scientific += 1
 974                self._advance()
 975            elif self._peek.upper() == "E" and not scientific:  # type: ignore
 976                scientific += 1
 977                self._advance()
 978            elif self._peek.isidentifier():  # type: ignore
 979                number_text = self._text
 980                literal = []
 981
 982                while self._peek.strip() and self._peek not in self.SINGLE_TOKENS:  # type: ignore
 983                    literal.append(self._peek.upper())  # type: ignore
 984                    self._advance()
 985
 986                literal = "".join(literal)  # type: ignore
 987                token_type = self.KEYWORDS.get(self.NUMERIC_LITERALS.get(literal))  # type: ignore
 988
 989                if token_type:
 990                    self._add(TokenType.NUMBER, number_text)
 991                    self._add(TokenType.DCOLON, "::")
 992                    return self._add(token_type, literal)  # type: ignore
 993                elif self.IDENTIFIER_CAN_START_WITH_DIGIT:
 994                    return self._add(TokenType.VAR)
 995
 996                self._add(TokenType.NUMBER, number_text)
 997                return self._advance(-len(literal))
 998            else:
 999                return self._add(TokenType.NUMBER)
1000
1001    def _scan_bits(self) -> None:
1002        self._advance()
1003        value = self._extract_value()
1004        try:
1005            self._add(TokenType.BIT_STRING, f"{int(value, 2)}")
1006        except ValueError:
1007            self._add(TokenType.IDENTIFIER)
1008
1009    def _scan_hex(self) -> None:
1010        self._advance()
1011        value = self._extract_value()
1012        try:
1013            self._add(TokenType.HEX_STRING, f"{int(value, 16)}")
1014        except ValueError:
1015            self._add(TokenType.IDENTIFIER)
1016
1017    def _extract_value(self) -> str:
1018        while True:
1019            char = self._peek.strip()  # type: ignore
1020            if char and char not in self.SINGLE_TOKENS:
1021                self._advance()
1022            else:
1023                break
1024
1025        return self._text
1026
1027    def _scan_string(self, quote: str) -> bool:
1028        quote_end = self._QUOTES.get(quote)  # type: ignore
1029        if quote_end is None:
1030            return False
1031
1032        self._advance(len(quote))
1033        text = self._extract_string(quote_end)
1034        text = text.encode(self.ENCODE).decode(self.ENCODE) if self.ENCODE else text  # type: ignore
1035        self._add(TokenType.NATIONAL if quote[0].upper() == "N" else TokenType.STRING, text)
1036        return True
1037
1038    # X'1234, b'0110', E'\\\\\' etc.
1039    def _scan_formatted_string(self, string_start: str) -> bool:
1040        if string_start in self._HEX_STRINGS:  # type: ignore
1041            delimiters = self._HEX_STRINGS  # type: ignore
1042            token_type = TokenType.HEX_STRING
1043            base = 16
1044        elif string_start in self._BIT_STRINGS:  # type: ignore
1045            delimiters = self._BIT_STRINGS  # type: ignore
1046            token_type = TokenType.BIT_STRING
1047            base = 2
1048        elif string_start in self._BYTE_STRINGS:  # type: ignore
1049            delimiters = self._BYTE_STRINGS  # type: ignore
1050            token_type = TokenType.BYTE_STRING
1051            base = None
1052        else:
1053            return False
1054
1055        self._advance(len(string_start))
1056        string_end = delimiters.get(string_start)
1057        text = self._extract_string(string_end)
1058
1059        if base is None:
1060            self._add(token_type, text)
1061        else:
1062            try:
1063                self._add(token_type, f"{int(text, base)}")
1064            except:
1065                raise RuntimeError(
1066                    f"Numeric string contains invalid characters from {self._line}:{self._start}"
1067                )
1068
1069        return True
1070
1071    def _scan_identifier(self, identifier_end: str) -> None:
1072        text = ""
1073        identifier_end_is_escape = identifier_end in self._IDENTIFIER_ESCAPES
1074
1075        while True:
1076            if self._end:
1077                raise RuntimeError(f"Missing {identifier_end} from {self._line}:{self._start}")
1078
1079            self._advance()
1080            if self._char == identifier_end:
1081                if identifier_end_is_escape and self._peek == identifier_end:
1082                    text += identifier_end  # type: ignore
1083                    self._advance()
1084                    continue
1085
1086                break
1087
1088            text += self._char  # type: ignore
1089
1090        self._add(TokenType.IDENTIFIER, text)
1091
1092    def _scan_var(self) -> None:
1093        while True:
1094            char = self._peek.strip()  # type: ignore
1095            if char and char not in self.SINGLE_TOKENS:
1096                self._advance()
1097            else:
1098                break
1099        self._add(
1100            TokenType.VAR
1101            if self._prev_token_type == TokenType.PARAMETER
1102            else self.KEYWORDS.get(self._text.upper(), TokenType.VAR)
1103        )
1104
1105    def _extract_string(self, delimiter: str) -> str:
1106        text = ""
1107        delim_size = len(delimiter)
1108
1109        while True:
1110            if self._char in self._STRING_ESCAPES and (
1111                self._peek == delimiter or self._peek in self._STRING_ESCAPES
1112            ):
1113                if self._peek == delimiter:
1114                    text += self._peek  # type: ignore
1115                else:
1116                    text += self._char + self._peek  # type: ignore
1117
1118                if self._current + 1 < self.size:
1119                    self._advance(2)
1120                else:
1121                    raise RuntimeError(f"Missing {delimiter} from {self._line}:{self._current}")
1122            else:
1123                if self._chars(delim_size) == delimiter:
1124                    if delim_size > 1:
1125                        self._advance(delim_size - 1)
1126                    break
1127
1128                if self._end:
1129                    raise RuntimeError(f"Missing {delimiter} from {self._line}:{self._start}")
1130                text += self._char  # type: ignore
1131                self._advance()
1132
1133        return text