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