Edit on GitHub

sqlglot.dialects.bigquery

  1from __future__ import annotations
  2
  3import logging
  4import re
  5import typing as t
  6
  7from sqlglot import exp, generator, parser, tokens, transforms
  8from sqlglot.dialects.dialect import (
  9    Dialect,
 10    NormalizationStrategy,
 11    arg_max_or_min_no_count,
 12    binary_from_function,
 13    date_add_interval_sql,
 14    datestrtodate_sql,
 15    build_formatted_time,
 16    filter_array_using_unnest,
 17    if_sql,
 18    inline_array_unless_query,
 19    max_or_greatest,
 20    min_or_least,
 21    no_ilike_sql,
 22    build_date_delta_with_interval,
 23    regexp_replace_sql,
 24    rename_func,
 25    sha256_sql,
 26    timestrtotime_sql,
 27    ts_or_ds_add_cast,
 28    unit_to_var,
 29)
 30from sqlglot.helper import seq_get, split_num_words
 31from sqlglot.tokens import TokenType
 32
 33if t.TYPE_CHECKING:
 34    from sqlglot._typing import E, Lit
 35
 36logger = logging.getLogger("sqlglot")
 37
 38
 39def _derived_table_values_to_unnest(self: BigQuery.Generator, expression: exp.Values) -> str:
 40    if not expression.find_ancestor(exp.From, exp.Join):
 41        return self.values_sql(expression)
 42
 43    structs = []
 44    alias = expression.args.get("alias")
 45    for tup in expression.find_all(exp.Tuple):
 46        field_aliases = (
 47            alias.columns
 48            if alias and alias.columns
 49            else (f"_c{i}" for i in range(len(tup.expressions)))
 50        )
 51        expressions = [
 52            exp.PropertyEQ(this=exp.to_identifier(name), expression=fld)
 53            for name, fld in zip(field_aliases, tup.expressions)
 54        ]
 55        structs.append(exp.Struct(expressions=expressions))
 56
 57    # Due to `UNNEST_COLUMN_ONLY`, it is expected that the table alias be contained in the columns expression
 58    alias_name_only = exp.TableAlias(columns=[alias.this]) if alias else None
 59    return self.unnest_sql(
 60        exp.Unnest(expressions=[exp.array(*structs, copy=False)], alias=alias_name_only)
 61    )
 62
 63
 64def _returnsproperty_sql(self: BigQuery.Generator, expression: exp.ReturnsProperty) -> str:
 65    this = expression.this
 66    if isinstance(this, exp.Schema):
 67        this = f"{self.sql(this, 'this')} <{self.expressions(this)}>"
 68    else:
 69        this = self.sql(this)
 70    return f"RETURNS {this}"
 71
 72
 73def _create_sql(self: BigQuery.Generator, expression: exp.Create) -> str:
 74    returns = expression.find(exp.ReturnsProperty)
 75    if expression.kind == "FUNCTION" and returns and returns.args.get("is_table"):
 76        expression.set("kind", "TABLE FUNCTION")
 77
 78        if isinstance(expression.expression, (exp.Subquery, exp.Literal)):
 79            expression.set("expression", expression.expression.this)
 80
 81    return self.create_sql(expression)
 82
 83
 84# https://issuetracker.google.com/issues/162294746
 85# workaround for bigquery bug when grouping by an expression and then ordering
 86# WITH x AS (SELECT 1 y)
 87# SELECT y + 1 z
 88# FROM x
 89# GROUP BY x + 1
 90# ORDER by z
 91def _alias_ordered_group(expression: exp.Expression) -> exp.Expression:
 92    if isinstance(expression, exp.Select):
 93        group = expression.args.get("group")
 94        order = expression.args.get("order")
 95
 96        if group and order:
 97            aliases = {
 98                select.this: select.args["alias"]
 99                for select in expression.selects
100                if isinstance(select, exp.Alias)
101            }
102
103            for grouped in group.expressions:
104                if grouped.is_int:
105                    continue
106                alias = aliases.get(grouped)
107                if alias:
108                    grouped.replace(exp.column(alias))
109
110    return expression
111
112
113def _pushdown_cte_column_names(expression: exp.Expression) -> exp.Expression:
114    """BigQuery doesn't allow column names when defining a CTE, so we try to push them down."""
115    if isinstance(expression, exp.CTE) and expression.alias_column_names:
116        cte_query = expression.this
117
118        if cte_query.is_star:
119            logger.warning(
120                "Can't push down CTE column names for star queries. Run the query through"
121                " the optimizer or use 'qualify' to expand the star projections first."
122            )
123            return expression
124
125        column_names = expression.alias_column_names
126        expression.args["alias"].set("columns", None)
127
128        for name, select in zip(column_names, cte_query.selects):
129            to_replace = select
130
131            if isinstance(select, exp.Alias):
132                select = select.this
133
134            # Inner aliases are shadowed by the CTE column names
135            to_replace.replace(exp.alias_(select, name))
136
137    return expression
138
139
140def _build_parse_timestamp(args: t.List) -> exp.StrToTime:
141    this = build_formatted_time(exp.StrToTime, "bigquery")([seq_get(args, 1), seq_get(args, 0)])
142    this.set("zone", seq_get(args, 2))
143    return this
144
145
146def _build_timestamp(args: t.List) -> exp.Timestamp:
147    timestamp = exp.Timestamp.from_arg_list(args)
148    timestamp.set("with_tz", True)
149    return timestamp
150
151
152def _build_date(args: t.List) -> exp.Date | exp.DateFromParts:
153    expr_type = exp.DateFromParts if len(args) == 3 else exp.Date
154    return expr_type.from_arg_list(args)
155
156
157def _build_to_hex(args: t.List) -> exp.Hex | exp.MD5:
158    # TO_HEX(MD5(..)) is common in BigQuery, so it's parsed into MD5 to simplify its transpilation
159    arg = seq_get(args, 0)
160    return exp.MD5(this=arg.this) if isinstance(arg, exp.MD5Digest) else exp.LowerHex(this=arg)
161
162
163def _array_contains_sql(self: BigQuery.Generator, expression: exp.ArrayContains) -> str:
164    return self.sql(
165        exp.Exists(
166            this=exp.select("1")
167            .from_(exp.Unnest(expressions=[expression.left]).as_("_unnest", table=["_col"]))
168            .where(exp.column("_col").eq(expression.right))
169        )
170    )
171
172
173def _ts_or_ds_add_sql(self: BigQuery.Generator, expression: exp.TsOrDsAdd) -> str:
174    return date_add_interval_sql("DATE", "ADD")(self, ts_or_ds_add_cast(expression))
175
176
177def _ts_or_ds_diff_sql(self: BigQuery.Generator, expression: exp.TsOrDsDiff) -> str:
178    expression.this.replace(exp.cast(expression.this, exp.DataType.Type.TIMESTAMP))
179    expression.expression.replace(exp.cast(expression.expression, exp.DataType.Type.TIMESTAMP))
180    unit = unit_to_var(expression)
181    return self.func("DATE_DIFF", expression.this, expression.expression, unit)
182
183
184def _unix_to_time_sql(self: BigQuery.Generator, expression: exp.UnixToTime) -> str:
185    scale = expression.args.get("scale")
186    timestamp = expression.this
187
188    if scale in (None, exp.UnixToTime.SECONDS):
189        return self.func("TIMESTAMP_SECONDS", timestamp)
190    if scale == exp.UnixToTime.MILLIS:
191        return self.func("TIMESTAMP_MILLIS", timestamp)
192    if scale == exp.UnixToTime.MICROS:
193        return self.func("TIMESTAMP_MICROS", timestamp)
194
195    unix_seconds = exp.cast(
196        exp.Div(this=timestamp, expression=exp.func("POW", 10, scale)), exp.DataType.Type.BIGINT
197    )
198    return self.func("TIMESTAMP_SECONDS", unix_seconds)
199
200
201def _build_time(args: t.List) -> exp.Func:
202    if len(args) == 1:
203        return exp.TsOrDsToTime(this=args[0])
204    if len(args) == 2:
205        return exp.Time.from_arg_list(args)
206    return exp.TimeFromParts.from_arg_list(args)
207
208
209def _build_datetime(args: t.List) -> exp.Func:
210    if len(args) == 1:
211        return exp.TsOrDsToTimestamp.from_arg_list(args)
212    if len(args) == 2:
213        return exp.Datetime.from_arg_list(args)
214    return exp.TimestampFromParts.from_arg_list(args)
215
216
217class BigQuery(Dialect):
218    WEEK_OFFSET = -1
219    UNNEST_COLUMN_ONLY = True
220    SUPPORTS_USER_DEFINED_TYPES = False
221    SUPPORTS_SEMI_ANTI_JOIN = False
222    LOG_BASE_FIRST = False
223    HEX_LOWERCASE = True
224
225    # https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#case_sensitivity
226    NORMALIZATION_STRATEGY = NormalizationStrategy.CASE_INSENSITIVE
227
228    # bigquery udfs are case sensitive
229    NORMALIZE_FUNCTIONS = False
230
231    # https://cloud.google.com/bigquery/docs/reference/standard-sql/format-elements#format_elements_date_time
232    TIME_MAPPING = {
233        "%D": "%m/%d/%y",
234        "%E6S": "%S.%f",
235    }
236
237    FORMAT_MAPPING = {
238        "DD": "%d",
239        "MM": "%m",
240        "MON": "%b",
241        "MONTH": "%B",
242        "YYYY": "%Y",
243        "YY": "%y",
244        "HH": "%I",
245        "HH12": "%I",
246        "HH24": "%H",
247        "MI": "%M",
248        "SS": "%S",
249        "SSSSS": "%f",
250        "TZH": "%z",
251    }
252
253    # The _PARTITIONTIME and _PARTITIONDATE pseudo-columns are not returned by a SELECT * statement
254    # https://cloud.google.com/bigquery/docs/querying-partitioned-tables#query_an_ingestion-time_partitioned_table
255    PSEUDOCOLUMNS = {"_PARTITIONTIME", "_PARTITIONDATE"}
256
257    def normalize_identifier(self, expression: E) -> E:
258        if (
259            isinstance(expression, exp.Identifier)
260            and self.normalization_strategy is not NormalizationStrategy.CASE_SENSITIVE
261        ):
262            parent = expression.parent
263            while isinstance(parent, exp.Dot):
264                parent = parent.parent
265
266            # In BigQuery, CTEs are case-insensitive, but UDF and table names are case-sensitive
267            # by default. The following check uses a heuristic to detect tables based on whether
268            # they are qualified. This should generally be correct, because tables in BigQuery
269            # must be qualified with at least a dataset, unless @@dataset_id is set.
270            case_sensitive = (
271                isinstance(parent, exp.UserDefinedFunction)
272                or (
273                    isinstance(parent, exp.Table)
274                    and parent.db
275                    and (parent.meta.get("quoted_table") or not parent.meta.get("maybe_column"))
276                )
277                or expression.meta.get("is_table")
278            )
279            if not case_sensitive:
280                expression.set("this", expression.this.lower())
281
282        return expression
283
284    class Tokenizer(tokens.Tokenizer):
285        QUOTES = ["'", '"', '"""', "'''"]
286        COMMENTS = ["--", "#", ("/*", "*/")]
287        IDENTIFIERS = ["`"]
288        STRING_ESCAPES = ["\\"]
289
290        HEX_STRINGS = [("0x", ""), ("0X", "")]
291
292        BYTE_STRINGS = [
293            (prefix + q, q) for q in t.cast(t.List[str], QUOTES) for prefix in ("b", "B")
294        ]
295
296        RAW_STRINGS = [
297            (prefix + q, q) for q in t.cast(t.List[str], QUOTES) for prefix in ("r", "R")
298        ]
299
300        KEYWORDS = {
301            **tokens.Tokenizer.KEYWORDS,
302            "ANY TYPE": TokenType.VARIANT,
303            "BEGIN": TokenType.COMMAND,
304            "BEGIN TRANSACTION": TokenType.BEGIN,
305            "BYTES": TokenType.BINARY,
306            "CURRENT_DATETIME": TokenType.CURRENT_DATETIME,
307            "DATETIME": TokenType.TIMESTAMP,
308            "DECLARE": TokenType.COMMAND,
309            "ELSEIF": TokenType.COMMAND,
310            "EXCEPTION": TokenType.COMMAND,
311            "FLOAT64": TokenType.DOUBLE,
312            "FOR SYSTEM_TIME": TokenType.TIMESTAMP_SNAPSHOT,
313            "MODEL": TokenType.MODEL,
314            "NOT DETERMINISTIC": TokenType.VOLATILE,
315            "RECORD": TokenType.STRUCT,
316            "TIMESTAMP": TokenType.TIMESTAMPTZ,
317        }
318        KEYWORDS.pop("DIV")
319        KEYWORDS.pop("VALUES")
320
321    class Parser(parser.Parser):
322        PREFIXED_PIVOT_COLUMNS = True
323        LOG_DEFAULTS_TO_LN = True
324        SUPPORTS_IMPLICIT_UNNEST = True
325
326        FUNCTIONS = {
327            **parser.Parser.FUNCTIONS,
328            "DATE": _build_date,
329            "DATE_ADD": build_date_delta_with_interval(exp.DateAdd),
330            "DATE_SUB": build_date_delta_with_interval(exp.DateSub),
331            "DATE_TRUNC": lambda args: exp.DateTrunc(
332                unit=exp.Literal.string(str(seq_get(args, 1))),
333                this=seq_get(args, 0),
334            ),
335            "DATETIME": _build_datetime,
336            "DATETIME_ADD": build_date_delta_with_interval(exp.DatetimeAdd),
337            "DATETIME_SUB": build_date_delta_with_interval(exp.DatetimeSub),
338            "DIV": binary_from_function(exp.IntDiv),
339            "FORMAT_DATE": lambda args: exp.TimeToStr(
340                this=exp.TsOrDsToDate(this=seq_get(args, 1)), format=seq_get(args, 0)
341            ),
342            "GENERATE_ARRAY": exp.GenerateSeries.from_arg_list,
343            "JSON_EXTRACT_SCALAR": lambda args: exp.JSONExtractScalar(
344                this=seq_get(args, 0), expression=seq_get(args, 1) or exp.Literal.string("$")
345            ),
346            "MD5": exp.MD5Digest.from_arg_list,
347            "TO_HEX": _build_to_hex,
348            "PARSE_DATE": lambda args: build_formatted_time(exp.StrToDate, "bigquery")(
349                [seq_get(args, 1), seq_get(args, 0)]
350            ),
351            "PARSE_TIMESTAMP": _build_parse_timestamp,
352            "REGEXP_CONTAINS": exp.RegexpLike.from_arg_list,
353            "REGEXP_EXTRACT": lambda args: exp.RegexpExtract(
354                this=seq_get(args, 0),
355                expression=seq_get(args, 1),
356                position=seq_get(args, 2),
357                occurrence=seq_get(args, 3),
358                group=exp.Literal.number(1) if re.compile(args[1].name).groups == 1 else None,
359            ),
360            "SHA256": lambda args: exp.SHA2(this=seq_get(args, 0), length=exp.Literal.number(256)),
361            "SHA512": lambda args: exp.SHA2(this=seq_get(args, 0), length=exp.Literal.number(512)),
362            "SPLIT": lambda args: exp.Split(
363                # https://cloud.google.com/bigquery/docs/reference/standard-sql/string_functions#split
364                this=seq_get(args, 0),
365                expression=seq_get(args, 1) or exp.Literal.string(","),
366            ),
367            "TIME": _build_time,
368            "TIME_ADD": build_date_delta_with_interval(exp.TimeAdd),
369            "TIME_SUB": build_date_delta_with_interval(exp.TimeSub),
370            "TIMESTAMP": _build_timestamp,
371            "TIMESTAMP_ADD": build_date_delta_with_interval(exp.TimestampAdd),
372            "TIMESTAMP_SUB": build_date_delta_with_interval(exp.TimestampSub),
373            "TIMESTAMP_MICROS": lambda args: exp.UnixToTime(
374                this=seq_get(args, 0), scale=exp.UnixToTime.MICROS
375            ),
376            "TIMESTAMP_MILLIS": lambda args: exp.UnixToTime(
377                this=seq_get(args, 0), scale=exp.UnixToTime.MILLIS
378            ),
379            "TIMESTAMP_SECONDS": lambda args: exp.UnixToTime(this=seq_get(args, 0)),
380            "TO_JSON_STRING": exp.JSONFormat.from_arg_list,
381        }
382
383        FUNCTION_PARSERS = {
384            **parser.Parser.FUNCTION_PARSERS,
385            "ARRAY": lambda self: self.expression(exp.Array, expressions=[self._parse_statement()]),
386        }
387        FUNCTION_PARSERS.pop("TRIM")
388
389        NO_PAREN_FUNCTIONS = {
390            **parser.Parser.NO_PAREN_FUNCTIONS,
391            TokenType.CURRENT_DATETIME: exp.CurrentDatetime,
392        }
393
394        NESTED_TYPE_TOKENS = {
395            *parser.Parser.NESTED_TYPE_TOKENS,
396            TokenType.TABLE,
397        }
398
399        PROPERTY_PARSERS = {
400            **parser.Parser.PROPERTY_PARSERS,
401            "NOT DETERMINISTIC": lambda self: self.expression(
402                exp.StabilityProperty, this=exp.Literal.string("VOLATILE")
403            ),
404            "OPTIONS": lambda self: self._parse_with_property(),
405        }
406
407        CONSTRAINT_PARSERS = {
408            **parser.Parser.CONSTRAINT_PARSERS,
409            "OPTIONS": lambda self: exp.Properties(expressions=self._parse_with_property()),
410        }
411
412        RANGE_PARSERS = parser.Parser.RANGE_PARSERS.copy()
413        RANGE_PARSERS.pop(TokenType.OVERLAPS)
414
415        NULL_TOKENS = {TokenType.NULL, TokenType.UNKNOWN}
416
417        STATEMENT_PARSERS = {
418            **parser.Parser.STATEMENT_PARSERS,
419            TokenType.ELSE: lambda self: self._parse_as_command(self._prev),
420            TokenType.END: lambda self: self._parse_as_command(self._prev),
421            TokenType.FOR: lambda self: self._parse_for_in(),
422        }
423
424        BRACKET_OFFSETS = {
425            "OFFSET": (0, False),
426            "ORDINAL": (1, False),
427            "SAFE_OFFSET": (0, True),
428            "SAFE_ORDINAL": (1, True),
429        }
430
431        def _parse_for_in(self) -> exp.ForIn:
432            this = self._parse_range()
433            self._match_text_seq("DO")
434            return self.expression(exp.ForIn, this=this, expression=self._parse_statement())
435
436        def _parse_table_part(self, schema: bool = False) -> t.Optional[exp.Expression]:
437            this = super()._parse_table_part(schema=schema) or self._parse_number()
438
439            # https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#table_names
440            if isinstance(this, exp.Identifier):
441                table_name = this.name
442                while self._match(TokenType.DASH, advance=False) and self._next:
443                    text = ""
444                    while self._curr and self._curr.token_type != TokenType.DOT:
445                        self._advance()
446                        text += self._prev.text
447                    table_name += text
448
449                this = exp.Identifier(this=table_name, quoted=this.args.get("quoted"))
450            elif isinstance(this, exp.Literal):
451                table_name = this.name
452
453                if self._is_connected() and self._parse_var(any_token=True):
454                    table_name += self._prev.text
455
456                this = exp.Identifier(this=table_name, quoted=True)
457
458            return this
459
460        def _parse_table_parts(
461            self, schema: bool = False, is_db_reference: bool = False, wildcard: bool = False
462        ) -> exp.Table:
463            table = super()._parse_table_parts(
464                schema=schema, is_db_reference=is_db_reference, wildcard=True
465            )
466
467            # proj-1.db.tbl -- `1.` is tokenized as a float so we need to unravel it here
468            if not table.catalog:
469                if table.db:
470                    parts = table.db.split(".")
471                    if len(parts) == 2 and not table.args["db"].quoted:
472                        table.set("catalog", exp.Identifier(this=parts[0]))
473                        table.set("db", exp.Identifier(this=parts[1]))
474                else:
475                    parts = table.name.split(".")
476                    if len(parts) == 2 and not table.this.quoted:
477                        table.set("db", exp.Identifier(this=parts[0]))
478                        table.set("this", exp.Identifier(this=parts[1]))
479
480            if any("." in p.name for p in table.parts):
481                catalog, db, this, *rest = (
482                    exp.to_identifier(p, quoted=True)
483                    for p in split_num_words(".".join(p.name for p in table.parts), ".", 3)
484                )
485
486                if rest and this:
487                    this = exp.Dot.build([this, *rest])  # type: ignore
488
489                table = exp.Table(
490                    this=this, db=db, catalog=catalog, pivots=table.args.get("pivots")
491                )
492                table.meta["quoted_table"] = True
493
494            return table
495
496        def _parse_column(self) -> t.Optional[exp.Expression]:
497            column = super()._parse_column()
498            if isinstance(column, exp.Column):
499                parts = column.parts
500                if any("." in p.name for p in parts):
501                    catalog, db, table, this, *rest = (
502                        exp.to_identifier(p, quoted=True)
503                        for p in split_num_words(".".join(p.name for p in parts), ".", 4)
504                    )
505
506                    if rest and this:
507                        this = exp.Dot.build([this, *rest])  # type: ignore
508
509                    column = exp.Column(this=this, table=table, db=db, catalog=catalog)
510                    column.meta["quoted_column"] = True
511
512            return column
513
514        @t.overload
515        def _parse_json_object(self, agg: Lit[False]) -> exp.JSONObject: ...
516
517        @t.overload
518        def _parse_json_object(self, agg: Lit[True]) -> exp.JSONObjectAgg: ...
519
520        def _parse_json_object(self, agg=False):
521            json_object = super()._parse_json_object()
522            array_kv_pair = seq_get(json_object.expressions, 0)
523
524            # Converts BQ's "signature 2" of JSON_OBJECT into SQLGlot's canonical representation
525            # https://cloud.google.com/bigquery/docs/reference/standard-sql/json_functions#json_object_signature2
526            if (
527                array_kv_pair
528                and isinstance(array_kv_pair.this, exp.Array)
529                and isinstance(array_kv_pair.expression, exp.Array)
530            ):
531                keys = array_kv_pair.this.expressions
532                values = array_kv_pair.expression.expressions
533
534                json_object.set(
535                    "expressions",
536                    [exp.JSONKeyValue(this=k, expression=v) for k, v in zip(keys, values)],
537                )
538
539            return json_object
540
541        def _parse_bracket(
542            self, this: t.Optional[exp.Expression] = None
543        ) -> t.Optional[exp.Expression]:
544            bracket = super()._parse_bracket(this)
545
546            if this is bracket:
547                return bracket
548
549            if isinstance(bracket, exp.Bracket):
550                for expression in bracket.expressions:
551                    name = expression.name.upper()
552
553                    if name not in self.BRACKET_OFFSETS:
554                        break
555
556                    offset, safe = self.BRACKET_OFFSETS[name]
557                    bracket.set("offset", offset)
558                    bracket.set("safe", safe)
559                    expression.replace(expression.expressions[0])
560
561            return bracket
562
563    class Generator(generator.Generator):
564        EXPLICIT_SET_OP = True
565        INTERVAL_ALLOWS_PLURAL_FORM = False
566        JOIN_HINTS = False
567        QUERY_HINTS = False
568        TABLE_HINTS = False
569        LIMIT_FETCH = "LIMIT"
570        RENAME_TABLE_WITH_DB = False
571        NVL2_SUPPORTED = False
572        UNNEST_WITH_ORDINALITY = False
573        COLLATE_IS_FUNC = True
574        LIMIT_ONLY_LITERALS = True
575        SUPPORTS_TABLE_ALIAS_COLUMNS = False
576        UNPIVOT_ALIASES_ARE_IDENTIFIERS = False
577        JSON_KEY_VALUE_PAIR_SEP = ","
578        NULL_ORDERING_SUPPORTED = False
579        IGNORE_NULLS_IN_FUNC = True
580        JSON_PATH_SINGLE_QUOTE_ESCAPE = True
581        CAN_IMPLEMENT_ARRAY_ANY = True
582        SUPPORTS_TO_NUMBER = False
583        NAMED_PLACEHOLDER_TOKEN = "@"
584        HEX_FUNC = "TO_HEX"
585        WITH_PROPERTIES_PREFIX = "OPTIONS"
586
587        TRANSFORMS = {
588            **generator.Generator.TRANSFORMS,
589            exp.ApproxDistinct: rename_func("APPROX_COUNT_DISTINCT"),
590            exp.ArgMax: arg_max_or_min_no_count("MAX_BY"),
591            exp.ArgMin: arg_max_or_min_no_count("MIN_BY"),
592            exp.Array: inline_array_unless_query,
593            exp.ArrayContains: _array_contains_sql,
594            exp.ArrayFilter: filter_array_using_unnest,
595            exp.ArraySize: rename_func("ARRAY_LENGTH"),
596            exp.Cast: transforms.preprocess([transforms.remove_precision_parameterized_types]),
597            exp.CollateProperty: lambda self, e: (
598                f"DEFAULT COLLATE {self.sql(e, 'this')}"
599                if e.args.get("default")
600                else f"COLLATE {self.sql(e, 'this')}"
601            ),
602            exp.Commit: lambda *_: "COMMIT TRANSACTION",
603            exp.CountIf: rename_func("COUNTIF"),
604            exp.Create: _create_sql,
605            exp.CTE: transforms.preprocess([_pushdown_cte_column_names]),
606            exp.DateAdd: date_add_interval_sql("DATE", "ADD"),
607            exp.DateDiff: lambda self, e: self.func(
608                "DATE_DIFF", e.this, e.expression, unit_to_var(e)
609            ),
610            exp.DateFromParts: rename_func("DATE"),
611            exp.DateStrToDate: datestrtodate_sql,
612            exp.DateSub: date_add_interval_sql("DATE", "SUB"),
613            exp.DatetimeAdd: date_add_interval_sql("DATETIME", "ADD"),
614            exp.DatetimeSub: date_add_interval_sql("DATETIME", "SUB"),
615            exp.DateTrunc: lambda self, e: self.func("DATE_TRUNC", e.this, e.text("unit")),
616            exp.FromTimeZone: lambda self, e: self.func(
617                "DATETIME", self.func("TIMESTAMP", e.this, e.args.get("zone")), "'UTC'"
618            ),
619            exp.GenerateSeries: rename_func("GENERATE_ARRAY"),
620            exp.GroupConcat: rename_func("STRING_AGG"),
621            exp.Hex: lambda self, e: self.func("UPPER", self.func("TO_HEX", self.sql(e, "this"))),
622            exp.If: if_sql(false_value="NULL"),
623            exp.ILike: no_ilike_sql,
624            exp.IntDiv: rename_func("DIV"),
625            exp.JSONFormat: rename_func("TO_JSON_STRING"),
626            exp.Max: max_or_greatest,
627            exp.MD5: lambda self, e: self.func("TO_HEX", self.func("MD5", e.this)),
628            exp.MD5Digest: rename_func("MD5"),
629            exp.Min: min_or_least,
630            exp.PartitionedByProperty: lambda self, e: f"PARTITION BY {self.sql(e, 'this')}",
631            exp.RegexpExtract: lambda self, e: self.func(
632                "REGEXP_EXTRACT",
633                e.this,
634                e.expression,
635                e.args.get("position"),
636                e.args.get("occurrence"),
637            ),
638            exp.RegexpReplace: regexp_replace_sql,
639            exp.RegexpLike: rename_func("REGEXP_CONTAINS"),
640            exp.ReturnsProperty: _returnsproperty_sql,
641            exp.Rollback: lambda *_: "ROLLBACK TRANSACTION",
642            exp.Select: transforms.preprocess(
643                [
644                    transforms.explode_to_unnest(),
645                    transforms.unqualify_unnest,
646                    transforms.eliminate_distinct_on,
647                    _alias_ordered_group,
648                    transforms.eliminate_semi_and_anti_joins,
649                ]
650            ),
651            exp.SHA: rename_func("SHA1"),
652            exp.SHA2: sha256_sql,
653            exp.StabilityProperty: lambda self, e: (
654                "DETERMINISTIC" if e.name == "IMMUTABLE" else "NOT DETERMINISTIC"
655            ),
656            exp.StrToDate: lambda self, e: self.func("PARSE_DATE", self.format_time(e), e.this),
657            exp.StrToTime: lambda self, e: self.func(
658                "PARSE_TIMESTAMP", self.format_time(e), e.this, e.args.get("zone")
659            ),
660            exp.TimeAdd: date_add_interval_sql("TIME", "ADD"),
661            exp.TimeFromParts: rename_func("TIME"),
662            exp.TimestampFromParts: rename_func("DATETIME"),
663            exp.TimeSub: date_add_interval_sql("TIME", "SUB"),
664            exp.TimestampAdd: date_add_interval_sql("TIMESTAMP", "ADD"),
665            exp.TimestampDiff: rename_func("TIMESTAMP_DIFF"),
666            exp.TimestampSub: date_add_interval_sql("TIMESTAMP", "SUB"),
667            exp.TimeStrToTime: timestrtotime_sql,
668            exp.Transaction: lambda *_: "BEGIN TRANSACTION",
669            exp.Trim: lambda self, e: self.func("TRIM", e.this, e.expression),
670            exp.TsOrDsAdd: _ts_or_ds_add_sql,
671            exp.TsOrDsDiff: _ts_or_ds_diff_sql,
672            exp.TsOrDsToTime: rename_func("TIME"),
673            exp.TsOrDsToTimestamp: rename_func("DATETIME"),
674            exp.Unhex: rename_func("FROM_HEX"),
675            exp.UnixDate: rename_func("UNIX_DATE"),
676            exp.UnixToTime: _unix_to_time_sql,
677            exp.Values: _derived_table_values_to_unnest,
678            exp.VariancePop: rename_func("VAR_POP"),
679        }
680
681        SUPPORTED_JSON_PATH_PARTS = {
682            exp.JSONPathKey,
683            exp.JSONPathRoot,
684            exp.JSONPathSubscript,
685        }
686
687        TYPE_MAPPING = {
688            **generator.Generator.TYPE_MAPPING,
689            exp.DataType.Type.BIGDECIMAL: "BIGNUMERIC",
690            exp.DataType.Type.BIGINT: "INT64",
691            exp.DataType.Type.BINARY: "BYTES",
692            exp.DataType.Type.BOOLEAN: "BOOL",
693            exp.DataType.Type.CHAR: "STRING",
694            exp.DataType.Type.DECIMAL: "NUMERIC",
695            exp.DataType.Type.DOUBLE: "FLOAT64",
696            exp.DataType.Type.FLOAT: "FLOAT64",
697            exp.DataType.Type.INT: "INT64",
698            exp.DataType.Type.NCHAR: "STRING",
699            exp.DataType.Type.NVARCHAR: "STRING",
700            exp.DataType.Type.SMALLINT: "INT64",
701            exp.DataType.Type.TEXT: "STRING",
702            exp.DataType.Type.TIMESTAMP: "DATETIME",
703            exp.DataType.Type.TIMESTAMPTZ: "TIMESTAMP",
704            exp.DataType.Type.TIMESTAMPLTZ: "TIMESTAMP",
705            exp.DataType.Type.TINYINT: "INT64",
706            exp.DataType.Type.VARBINARY: "BYTES",
707            exp.DataType.Type.ROWVERSION: "BYTES",
708            exp.DataType.Type.VARCHAR: "STRING",
709            exp.DataType.Type.VARIANT: "ANY TYPE",
710        }
711
712        PROPERTIES_LOCATION = {
713            **generator.Generator.PROPERTIES_LOCATION,
714            exp.PartitionedByProperty: exp.Properties.Location.POST_SCHEMA,
715            exp.VolatileProperty: exp.Properties.Location.UNSUPPORTED,
716        }
717
718        # from: https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#reserved_keywords
719        RESERVED_KEYWORDS = {
720            "all",
721            "and",
722            "any",
723            "array",
724            "as",
725            "asc",
726            "assert_rows_modified",
727            "at",
728            "between",
729            "by",
730            "case",
731            "cast",
732            "collate",
733            "contains",
734            "create",
735            "cross",
736            "cube",
737            "current",
738            "default",
739            "define",
740            "desc",
741            "distinct",
742            "else",
743            "end",
744            "enum",
745            "escape",
746            "except",
747            "exclude",
748            "exists",
749            "extract",
750            "false",
751            "fetch",
752            "following",
753            "for",
754            "from",
755            "full",
756            "group",
757            "grouping",
758            "groups",
759            "hash",
760            "having",
761            "if",
762            "ignore",
763            "in",
764            "inner",
765            "intersect",
766            "interval",
767            "into",
768            "is",
769            "join",
770            "lateral",
771            "left",
772            "like",
773            "limit",
774            "lookup",
775            "merge",
776            "natural",
777            "new",
778            "no",
779            "not",
780            "null",
781            "nulls",
782            "of",
783            "on",
784            "or",
785            "order",
786            "outer",
787            "over",
788            "partition",
789            "preceding",
790            "proto",
791            "qualify",
792            "range",
793            "recursive",
794            "respect",
795            "right",
796            "rollup",
797            "rows",
798            "select",
799            "set",
800            "some",
801            "struct",
802            "tablesample",
803            "then",
804            "to",
805            "treat",
806            "true",
807            "unbounded",
808            "union",
809            "unnest",
810            "using",
811            "when",
812            "where",
813            "window",
814            "with",
815            "within",
816        }
817
818        def mod_sql(self, expression: exp.Mod) -> str:
819            this = expression.this
820            expr = expression.expression
821            return self.func(
822                "MOD",
823                this.unnest() if isinstance(this, exp.Paren) else this,
824                expr.unnest() if isinstance(expr, exp.Paren) else expr,
825            )
826
827        def column_parts(self, expression: exp.Column) -> str:
828            if expression.meta.get("quoted_column"):
829                # If a column reference is of the form `dataset.table`.name, we need
830                # to preserve the quoted table path, otherwise the reference breaks
831                table_parts = ".".join(p.name for p in expression.parts[:-1])
832                table_path = self.sql(exp.Identifier(this=table_parts, quoted=True))
833                return f"{table_path}.{self.sql(expression, 'this')}"
834
835            return super().column_parts(expression)
836
837        def table_parts(self, expression: exp.Table) -> str:
838            # Depending on the context, `x.y` may not resolve to the same data source as `x`.`y`, so
839            # we need to make sure the correct quoting is used in each case.
840            #
841            # For example, if there is a CTE x that clashes with a schema name, then the former will
842            # return the table y in that schema, whereas the latter will return the CTE's y column:
843            #
844            # - WITH x AS (SELECT [1, 2] AS y) SELECT * FROM x, `x.y`   -> cross join
845            # - WITH x AS (SELECT [1, 2] AS y) SELECT * FROM x, `x`.`y` -> implicit unnest
846            if expression.meta.get("quoted_table"):
847                table_parts = ".".join(p.name for p in expression.parts)
848                return self.sql(exp.Identifier(this=table_parts, quoted=True))
849
850            return super().table_parts(expression)
851
852        def timetostr_sql(self, expression: exp.TimeToStr) -> str:
853            this = expression.this if isinstance(expression.this, exp.TsOrDsToDate) else expression
854            return self.func("FORMAT_DATE", self.format_time(expression), this.this)
855
856        def eq_sql(self, expression: exp.EQ) -> str:
857            # Operands of = cannot be NULL in BigQuery
858            if isinstance(expression.left, exp.Null) or isinstance(expression.right, exp.Null):
859                if not isinstance(expression.parent, exp.Update):
860                    return "NULL"
861
862            return self.binary(expression, "=")
863
864        def attimezone_sql(self, expression: exp.AtTimeZone) -> str:
865            parent = expression.parent
866
867            # BigQuery allows CAST(.. AS {STRING|TIMESTAMP} [FORMAT <fmt> [AT TIME ZONE <tz>]]).
868            # Only the TIMESTAMP one should use the below conversion, when AT TIME ZONE is included.
869            if not isinstance(parent, exp.Cast) or not parent.to.is_type("text"):
870                return self.func(
871                    "TIMESTAMP", self.func("DATETIME", expression.this, expression.args.get("zone"))
872                )
873
874            return super().attimezone_sql(expression)
875
876        def trycast_sql(self, expression: exp.TryCast) -> str:
877            return self.cast_sql(expression, safe_prefix="SAFE_")
878
879        def bracket_sql(self, expression: exp.Bracket) -> str:
880            this = expression.this
881            expressions = expression.expressions
882
883            if len(expressions) == 1 and this and this.is_type(exp.DataType.Type.STRUCT):
884                arg = expressions[0]
885                if arg.type is None:
886                    from sqlglot.optimizer.annotate_types import annotate_types
887
888                    arg = annotate_types(arg)
889
890                if arg.type and arg.type.this in exp.DataType.TEXT_TYPES:
891                    # BQ doesn't support bracket syntax with string values for structs
892                    return f"{self.sql(this)}.{arg.name}"
893
894            expressions_sql = self.expressions(expression, flat=True)
895            offset = expression.args.get("offset")
896
897            if offset == 0:
898                expressions_sql = f"OFFSET({expressions_sql})"
899            elif offset == 1:
900                expressions_sql = f"ORDINAL({expressions_sql})"
901            elif offset is not None:
902                self.unsupported(f"Unsupported array offset: {offset}")
903
904            if expression.args.get("safe"):
905                expressions_sql = f"SAFE_{expressions_sql}"
906
907            return f"{self.sql(this)}[{expressions_sql}]"
908
909        def in_unnest_op(self, expression: exp.Unnest) -> str:
910            return self.sql(expression)
911
912        def except_op(self, expression: exp.Except) -> str:
913            if not expression.args.get("distinct"):
914                self.unsupported("EXCEPT without DISTINCT is not supported in BigQuery")
915            return f"EXCEPT{' DISTINCT' if expression.args.get('distinct') else ' ALL'}"
916
917        def intersect_op(self, expression: exp.Intersect) -> str:
918            if not expression.args.get("distinct"):
919                self.unsupported("INTERSECT without DISTINCT is not supported in BigQuery")
920            return f"INTERSECT{' DISTINCT' if expression.args.get('distinct') else ' ALL'}"
921
922        def version_sql(self, expression: exp.Version) -> str:
923            if expression.name == "TIMESTAMP":
924                expression.set("this", "SYSTEM_TIME")
925            return super().version_sql(expression)
logger = <Logger sqlglot (WARNING)>
class BigQuery(sqlglot.dialects.dialect.Dialect):
218class BigQuery(Dialect):
219    WEEK_OFFSET = -1
220    UNNEST_COLUMN_ONLY = True
221    SUPPORTS_USER_DEFINED_TYPES = False
222    SUPPORTS_SEMI_ANTI_JOIN = False
223    LOG_BASE_FIRST = False
224    HEX_LOWERCASE = True
225
226    # https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#case_sensitivity
227    NORMALIZATION_STRATEGY = NormalizationStrategy.CASE_INSENSITIVE
228
229    # bigquery udfs are case sensitive
230    NORMALIZE_FUNCTIONS = False
231
232    # https://cloud.google.com/bigquery/docs/reference/standard-sql/format-elements#format_elements_date_time
233    TIME_MAPPING = {
234        "%D": "%m/%d/%y",
235        "%E6S": "%S.%f",
236    }
237
238    FORMAT_MAPPING = {
239        "DD": "%d",
240        "MM": "%m",
241        "MON": "%b",
242        "MONTH": "%B",
243        "YYYY": "%Y",
244        "YY": "%y",
245        "HH": "%I",
246        "HH12": "%I",
247        "HH24": "%H",
248        "MI": "%M",
249        "SS": "%S",
250        "SSSSS": "%f",
251        "TZH": "%z",
252    }
253
254    # The _PARTITIONTIME and _PARTITIONDATE pseudo-columns are not returned by a SELECT * statement
255    # https://cloud.google.com/bigquery/docs/querying-partitioned-tables#query_an_ingestion-time_partitioned_table
256    PSEUDOCOLUMNS = {"_PARTITIONTIME", "_PARTITIONDATE"}
257
258    def normalize_identifier(self, expression: E) -> E:
259        if (
260            isinstance(expression, exp.Identifier)
261            and self.normalization_strategy is not NormalizationStrategy.CASE_SENSITIVE
262        ):
263            parent = expression.parent
264            while isinstance(parent, exp.Dot):
265                parent = parent.parent
266
267            # In BigQuery, CTEs are case-insensitive, but UDF and table names are case-sensitive
268            # by default. The following check uses a heuristic to detect tables based on whether
269            # they are qualified. This should generally be correct, because tables in BigQuery
270            # must be qualified with at least a dataset, unless @@dataset_id is set.
271            case_sensitive = (
272                isinstance(parent, exp.UserDefinedFunction)
273                or (
274                    isinstance(parent, exp.Table)
275                    and parent.db
276                    and (parent.meta.get("quoted_table") or not parent.meta.get("maybe_column"))
277                )
278                or expression.meta.get("is_table")
279            )
280            if not case_sensitive:
281                expression.set("this", expression.this.lower())
282
283        return expression
284
285    class Tokenizer(tokens.Tokenizer):
286        QUOTES = ["'", '"', '"""', "'''"]
287        COMMENTS = ["--", "#", ("/*", "*/")]
288        IDENTIFIERS = ["`"]
289        STRING_ESCAPES = ["\\"]
290
291        HEX_STRINGS = [("0x", ""), ("0X", "")]
292
293        BYTE_STRINGS = [
294            (prefix + q, q) for q in t.cast(t.List[str], QUOTES) for prefix in ("b", "B")
295        ]
296
297        RAW_STRINGS = [
298            (prefix + q, q) for q in t.cast(t.List[str], QUOTES) for prefix in ("r", "R")
299        ]
300
301        KEYWORDS = {
302            **tokens.Tokenizer.KEYWORDS,
303            "ANY TYPE": TokenType.VARIANT,
304            "BEGIN": TokenType.COMMAND,
305            "BEGIN TRANSACTION": TokenType.BEGIN,
306            "BYTES": TokenType.BINARY,
307            "CURRENT_DATETIME": TokenType.CURRENT_DATETIME,
308            "DATETIME": TokenType.TIMESTAMP,
309            "DECLARE": TokenType.COMMAND,
310            "ELSEIF": TokenType.COMMAND,
311            "EXCEPTION": TokenType.COMMAND,
312            "FLOAT64": TokenType.DOUBLE,
313            "FOR SYSTEM_TIME": TokenType.TIMESTAMP_SNAPSHOT,
314            "MODEL": TokenType.MODEL,
315            "NOT DETERMINISTIC": TokenType.VOLATILE,
316            "RECORD": TokenType.STRUCT,
317            "TIMESTAMP": TokenType.TIMESTAMPTZ,
318        }
319        KEYWORDS.pop("DIV")
320        KEYWORDS.pop("VALUES")
321
322    class Parser(parser.Parser):
323        PREFIXED_PIVOT_COLUMNS = True
324        LOG_DEFAULTS_TO_LN = True
325        SUPPORTS_IMPLICIT_UNNEST = True
326
327        FUNCTIONS = {
328            **parser.Parser.FUNCTIONS,
329            "DATE": _build_date,
330            "DATE_ADD": build_date_delta_with_interval(exp.DateAdd),
331            "DATE_SUB": build_date_delta_with_interval(exp.DateSub),
332            "DATE_TRUNC": lambda args: exp.DateTrunc(
333                unit=exp.Literal.string(str(seq_get(args, 1))),
334                this=seq_get(args, 0),
335            ),
336            "DATETIME": _build_datetime,
337            "DATETIME_ADD": build_date_delta_with_interval(exp.DatetimeAdd),
338            "DATETIME_SUB": build_date_delta_with_interval(exp.DatetimeSub),
339            "DIV": binary_from_function(exp.IntDiv),
340            "FORMAT_DATE": lambda args: exp.TimeToStr(
341                this=exp.TsOrDsToDate(this=seq_get(args, 1)), format=seq_get(args, 0)
342            ),
343            "GENERATE_ARRAY": exp.GenerateSeries.from_arg_list,
344            "JSON_EXTRACT_SCALAR": lambda args: exp.JSONExtractScalar(
345                this=seq_get(args, 0), expression=seq_get(args, 1) or exp.Literal.string("$")
346            ),
347            "MD5": exp.MD5Digest.from_arg_list,
348            "TO_HEX": _build_to_hex,
349            "PARSE_DATE": lambda args: build_formatted_time(exp.StrToDate, "bigquery")(
350                [seq_get(args, 1), seq_get(args, 0)]
351            ),
352            "PARSE_TIMESTAMP": _build_parse_timestamp,
353            "REGEXP_CONTAINS": exp.RegexpLike.from_arg_list,
354            "REGEXP_EXTRACT": lambda args: exp.RegexpExtract(
355                this=seq_get(args, 0),
356                expression=seq_get(args, 1),
357                position=seq_get(args, 2),
358                occurrence=seq_get(args, 3),
359                group=exp.Literal.number(1) if re.compile(args[1].name).groups == 1 else None,
360            ),
361            "SHA256": lambda args: exp.SHA2(this=seq_get(args, 0), length=exp.Literal.number(256)),
362            "SHA512": lambda args: exp.SHA2(this=seq_get(args, 0), length=exp.Literal.number(512)),
363            "SPLIT": lambda args: exp.Split(
364                # https://cloud.google.com/bigquery/docs/reference/standard-sql/string_functions#split
365                this=seq_get(args, 0),
366                expression=seq_get(args, 1) or exp.Literal.string(","),
367            ),
368            "TIME": _build_time,
369            "TIME_ADD": build_date_delta_with_interval(exp.TimeAdd),
370            "TIME_SUB": build_date_delta_with_interval(exp.TimeSub),
371            "TIMESTAMP": _build_timestamp,
372            "TIMESTAMP_ADD": build_date_delta_with_interval(exp.TimestampAdd),
373            "TIMESTAMP_SUB": build_date_delta_with_interval(exp.TimestampSub),
374            "TIMESTAMP_MICROS": lambda args: exp.UnixToTime(
375                this=seq_get(args, 0), scale=exp.UnixToTime.MICROS
376            ),
377            "TIMESTAMP_MILLIS": lambda args: exp.UnixToTime(
378                this=seq_get(args, 0), scale=exp.UnixToTime.MILLIS
379            ),
380            "TIMESTAMP_SECONDS": lambda args: exp.UnixToTime(this=seq_get(args, 0)),
381            "TO_JSON_STRING": exp.JSONFormat.from_arg_list,
382        }
383
384        FUNCTION_PARSERS = {
385            **parser.Parser.FUNCTION_PARSERS,
386            "ARRAY": lambda self: self.expression(exp.Array, expressions=[self._parse_statement()]),
387        }
388        FUNCTION_PARSERS.pop("TRIM")
389
390        NO_PAREN_FUNCTIONS = {
391            **parser.Parser.NO_PAREN_FUNCTIONS,
392            TokenType.CURRENT_DATETIME: exp.CurrentDatetime,
393        }
394
395        NESTED_TYPE_TOKENS = {
396            *parser.Parser.NESTED_TYPE_TOKENS,
397            TokenType.TABLE,
398        }
399
400        PROPERTY_PARSERS = {
401            **parser.Parser.PROPERTY_PARSERS,
402            "NOT DETERMINISTIC": lambda self: self.expression(
403                exp.StabilityProperty, this=exp.Literal.string("VOLATILE")
404            ),
405            "OPTIONS": lambda self: self._parse_with_property(),
406        }
407
408        CONSTRAINT_PARSERS = {
409            **parser.Parser.CONSTRAINT_PARSERS,
410            "OPTIONS": lambda self: exp.Properties(expressions=self._parse_with_property()),
411        }
412
413        RANGE_PARSERS = parser.Parser.RANGE_PARSERS.copy()
414        RANGE_PARSERS.pop(TokenType.OVERLAPS)
415
416        NULL_TOKENS = {TokenType.NULL, TokenType.UNKNOWN}
417
418        STATEMENT_PARSERS = {
419            **parser.Parser.STATEMENT_PARSERS,
420            TokenType.ELSE: lambda self: self._parse_as_command(self._prev),
421            TokenType.END: lambda self: self._parse_as_command(self._prev),
422            TokenType.FOR: lambda self: self._parse_for_in(),
423        }
424
425        BRACKET_OFFSETS = {
426            "OFFSET": (0, False),
427            "ORDINAL": (1, False),
428            "SAFE_OFFSET": (0, True),
429            "SAFE_ORDINAL": (1, True),
430        }
431
432        def _parse_for_in(self) -> exp.ForIn:
433            this = self._parse_range()
434            self._match_text_seq("DO")
435            return self.expression(exp.ForIn, this=this, expression=self._parse_statement())
436
437        def _parse_table_part(self, schema: bool = False) -> t.Optional[exp.Expression]:
438            this = super()._parse_table_part(schema=schema) or self._parse_number()
439
440            # https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#table_names
441            if isinstance(this, exp.Identifier):
442                table_name = this.name
443                while self._match(TokenType.DASH, advance=False) and self._next:
444                    text = ""
445                    while self._curr and self._curr.token_type != TokenType.DOT:
446                        self._advance()
447                        text += self._prev.text
448                    table_name += text
449
450                this = exp.Identifier(this=table_name, quoted=this.args.get("quoted"))
451            elif isinstance(this, exp.Literal):
452                table_name = this.name
453
454                if self._is_connected() and self._parse_var(any_token=True):
455                    table_name += self._prev.text
456
457                this = exp.Identifier(this=table_name, quoted=True)
458
459            return this
460
461        def _parse_table_parts(
462            self, schema: bool = False, is_db_reference: bool = False, wildcard: bool = False
463        ) -> exp.Table:
464            table = super()._parse_table_parts(
465                schema=schema, is_db_reference=is_db_reference, wildcard=True
466            )
467
468            # proj-1.db.tbl -- `1.` is tokenized as a float so we need to unravel it here
469            if not table.catalog:
470                if table.db:
471                    parts = table.db.split(".")
472                    if len(parts) == 2 and not table.args["db"].quoted:
473                        table.set("catalog", exp.Identifier(this=parts[0]))
474                        table.set("db", exp.Identifier(this=parts[1]))
475                else:
476                    parts = table.name.split(".")
477                    if len(parts) == 2 and not table.this.quoted:
478                        table.set("db", exp.Identifier(this=parts[0]))
479                        table.set("this", exp.Identifier(this=parts[1]))
480
481            if any("." in p.name for p in table.parts):
482                catalog, db, this, *rest = (
483                    exp.to_identifier(p, quoted=True)
484                    for p in split_num_words(".".join(p.name for p in table.parts), ".", 3)
485                )
486
487                if rest and this:
488                    this = exp.Dot.build([this, *rest])  # type: ignore
489
490                table = exp.Table(
491                    this=this, db=db, catalog=catalog, pivots=table.args.get("pivots")
492                )
493                table.meta["quoted_table"] = True
494
495            return table
496
497        def _parse_column(self) -> t.Optional[exp.Expression]:
498            column = super()._parse_column()
499            if isinstance(column, exp.Column):
500                parts = column.parts
501                if any("." in p.name for p in parts):
502                    catalog, db, table, this, *rest = (
503                        exp.to_identifier(p, quoted=True)
504                        for p in split_num_words(".".join(p.name for p in parts), ".", 4)
505                    )
506
507                    if rest and this:
508                        this = exp.Dot.build([this, *rest])  # type: ignore
509
510                    column = exp.Column(this=this, table=table, db=db, catalog=catalog)
511                    column.meta["quoted_column"] = True
512
513            return column
514
515        @t.overload
516        def _parse_json_object(self, agg: Lit[False]) -> exp.JSONObject: ...
517
518        @t.overload
519        def _parse_json_object(self, agg: Lit[True]) -> exp.JSONObjectAgg: ...
520
521        def _parse_json_object(self, agg=False):
522            json_object = super()._parse_json_object()
523            array_kv_pair = seq_get(json_object.expressions, 0)
524
525            # Converts BQ's "signature 2" of JSON_OBJECT into SQLGlot's canonical representation
526            # https://cloud.google.com/bigquery/docs/reference/standard-sql/json_functions#json_object_signature2
527            if (
528                array_kv_pair
529                and isinstance(array_kv_pair.this, exp.Array)
530                and isinstance(array_kv_pair.expression, exp.Array)
531            ):
532                keys = array_kv_pair.this.expressions
533                values = array_kv_pair.expression.expressions
534
535                json_object.set(
536                    "expressions",
537                    [exp.JSONKeyValue(this=k, expression=v) for k, v in zip(keys, values)],
538                )
539
540            return json_object
541
542        def _parse_bracket(
543            self, this: t.Optional[exp.Expression] = None
544        ) -> t.Optional[exp.Expression]:
545            bracket = super()._parse_bracket(this)
546
547            if this is bracket:
548                return bracket
549
550            if isinstance(bracket, exp.Bracket):
551                for expression in bracket.expressions:
552                    name = expression.name.upper()
553
554                    if name not in self.BRACKET_OFFSETS:
555                        break
556
557                    offset, safe = self.BRACKET_OFFSETS[name]
558                    bracket.set("offset", offset)
559                    bracket.set("safe", safe)
560                    expression.replace(expression.expressions[0])
561
562            return bracket
563
564    class Generator(generator.Generator):
565        EXPLICIT_SET_OP = True
566        INTERVAL_ALLOWS_PLURAL_FORM = False
567        JOIN_HINTS = False
568        QUERY_HINTS = False
569        TABLE_HINTS = False
570        LIMIT_FETCH = "LIMIT"
571        RENAME_TABLE_WITH_DB = False
572        NVL2_SUPPORTED = False
573        UNNEST_WITH_ORDINALITY = False
574        COLLATE_IS_FUNC = True
575        LIMIT_ONLY_LITERALS = True
576        SUPPORTS_TABLE_ALIAS_COLUMNS = False
577        UNPIVOT_ALIASES_ARE_IDENTIFIERS = False
578        JSON_KEY_VALUE_PAIR_SEP = ","
579        NULL_ORDERING_SUPPORTED = False
580        IGNORE_NULLS_IN_FUNC = True
581        JSON_PATH_SINGLE_QUOTE_ESCAPE = True
582        CAN_IMPLEMENT_ARRAY_ANY = True
583        SUPPORTS_TO_NUMBER = False
584        NAMED_PLACEHOLDER_TOKEN = "@"
585        HEX_FUNC = "TO_HEX"
586        WITH_PROPERTIES_PREFIX = "OPTIONS"
587
588        TRANSFORMS = {
589            **generator.Generator.TRANSFORMS,
590            exp.ApproxDistinct: rename_func("APPROX_COUNT_DISTINCT"),
591            exp.ArgMax: arg_max_or_min_no_count("MAX_BY"),
592            exp.ArgMin: arg_max_or_min_no_count("MIN_BY"),
593            exp.Array: inline_array_unless_query,
594            exp.ArrayContains: _array_contains_sql,
595            exp.ArrayFilter: filter_array_using_unnest,
596            exp.ArraySize: rename_func("ARRAY_LENGTH"),
597            exp.Cast: transforms.preprocess([transforms.remove_precision_parameterized_types]),
598            exp.CollateProperty: lambda self, e: (
599                f"DEFAULT COLLATE {self.sql(e, 'this')}"
600                if e.args.get("default")
601                else f"COLLATE {self.sql(e, 'this')}"
602            ),
603            exp.Commit: lambda *_: "COMMIT TRANSACTION",
604            exp.CountIf: rename_func("COUNTIF"),
605            exp.Create: _create_sql,
606            exp.CTE: transforms.preprocess([_pushdown_cte_column_names]),
607            exp.DateAdd: date_add_interval_sql("DATE", "ADD"),
608            exp.DateDiff: lambda self, e: self.func(
609                "DATE_DIFF", e.this, e.expression, unit_to_var(e)
610            ),
611            exp.DateFromParts: rename_func("DATE"),
612            exp.DateStrToDate: datestrtodate_sql,
613            exp.DateSub: date_add_interval_sql("DATE", "SUB"),
614            exp.DatetimeAdd: date_add_interval_sql("DATETIME", "ADD"),
615            exp.DatetimeSub: date_add_interval_sql("DATETIME", "SUB"),
616            exp.DateTrunc: lambda self, e: self.func("DATE_TRUNC", e.this, e.text("unit")),
617            exp.FromTimeZone: lambda self, e: self.func(
618                "DATETIME", self.func("TIMESTAMP", e.this, e.args.get("zone")), "'UTC'"
619            ),
620            exp.GenerateSeries: rename_func("GENERATE_ARRAY"),
621            exp.GroupConcat: rename_func("STRING_AGG"),
622            exp.Hex: lambda self, e: self.func("UPPER", self.func("TO_HEX", self.sql(e, "this"))),
623            exp.If: if_sql(false_value="NULL"),
624            exp.ILike: no_ilike_sql,
625            exp.IntDiv: rename_func("DIV"),
626            exp.JSONFormat: rename_func("TO_JSON_STRING"),
627            exp.Max: max_or_greatest,
628            exp.MD5: lambda self, e: self.func("TO_HEX", self.func("MD5", e.this)),
629            exp.MD5Digest: rename_func("MD5"),
630            exp.Min: min_or_least,
631            exp.PartitionedByProperty: lambda self, e: f"PARTITION BY {self.sql(e, 'this')}",
632            exp.RegexpExtract: lambda self, e: self.func(
633                "REGEXP_EXTRACT",
634                e.this,
635                e.expression,
636                e.args.get("position"),
637                e.args.get("occurrence"),
638            ),
639            exp.RegexpReplace: regexp_replace_sql,
640            exp.RegexpLike: rename_func("REGEXP_CONTAINS"),
641            exp.ReturnsProperty: _returnsproperty_sql,
642            exp.Rollback: lambda *_: "ROLLBACK TRANSACTION",
643            exp.Select: transforms.preprocess(
644                [
645                    transforms.explode_to_unnest(),
646                    transforms.unqualify_unnest,
647                    transforms.eliminate_distinct_on,
648                    _alias_ordered_group,
649                    transforms.eliminate_semi_and_anti_joins,
650                ]
651            ),
652            exp.SHA: rename_func("SHA1"),
653            exp.SHA2: sha256_sql,
654            exp.StabilityProperty: lambda self, e: (
655                "DETERMINISTIC" if e.name == "IMMUTABLE" else "NOT DETERMINISTIC"
656            ),
657            exp.StrToDate: lambda self, e: self.func("PARSE_DATE", self.format_time(e), e.this),
658            exp.StrToTime: lambda self, e: self.func(
659                "PARSE_TIMESTAMP", self.format_time(e), e.this, e.args.get("zone")
660            ),
661            exp.TimeAdd: date_add_interval_sql("TIME", "ADD"),
662            exp.TimeFromParts: rename_func("TIME"),
663            exp.TimestampFromParts: rename_func("DATETIME"),
664            exp.TimeSub: date_add_interval_sql("TIME", "SUB"),
665            exp.TimestampAdd: date_add_interval_sql("TIMESTAMP", "ADD"),
666            exp.TimestampDiff: rename_func("TIMESTAMP_DIFF"),
667            exp.TimestampSub: date_add_interval_sql("TIMESTAMP", "SUB"),
668            exp.TimeStrToTime: timestrtotime_sql,
669            exp.Transaction: lambda *_: "BEGIN TRANSACTION",
670            exp.Trim: lambda self, e: self.func("TRIM", e.this, e.expression),
671            exp.TsOrDsAdd: _ts_or_ds_add_sql,
672            exp.TsOrDsDiff: _ts_or_ds_diff_sql,
673            exp.TsOrDsToTime: rename_func("TIME"),
674            exp.TsOrDsToTimestamp: rename_func("DATETIME"),
675            exp.Unhex: rename_func("FROM_HEX"),
676            exp.UnixDate: rename_func("UNIX_DATE"),
677            exp.UnixToTime: _unix_to_time_sql,
678            exp.Values: _derived_table_values_to_unnest,
679            exp.VariancePop: rename_func("VAR_POP"),
680        }
681
682        SUPPORTED_JSON_PATH_PARTS = {
683            exp.JSONPathKey,
684            exp.JSONPathRoot,
685            exp.JSONPathSubscript,
686        }
687
688        TYPE_MAPPING = {
689            **generator.Generator.TYPE_MAPPING,
690            exp.DataType.Type.BIGDECIMAL: "BIGNUMERIC",
691            exp.DataType.Type.BIGINT: "INT64",
692            exp.DataType.Type.BINARY: "BYTES",
693            exp.DataType.Type.BOOLEAN: "BOOL",
694            exp.DataType.Type.CHAR: "STRING",
695            exp.DataType.Type.DECIMAL: "NUMERIC",
696            exp.DataType.Type.DOUBLE: "FLOAT64",
697            exp.DataType.Type.FLOAT: "FLOAT64",
698            exp.DataType.Type.INT: "INT64",
699            exp.DataType.Type.NCHAR: "STRING",
700            exp.DataType.Type.NVARCHAR: "STRING",
701            exp.DataType.Type.SMALLINT: "INT64",
702            exp.DataType.Type.TEXT: "STRING",
703            exp.DataType.Type.TIMESTAMP: "DATETIME",
704            exp.DataType.Type.TIMESTAMPTZ: "TIMESTAMP",
705            exp.DataType.Type.TIMESTAMPLTZ: "TIMESTAMP",
706            exp.DataType.Type.TINYINT: "INT64",
707            exp.DataType.Type.VARBINARY: "BYTES",
708            exp.DataType.Type.ROWVERSION: "BYTES",
709            exp.DataType.Type.VARCHAR: "STRING",
710            exp.DataType.Type.VARIANT: "ANY TYPE",
711        }
712
713        PROPERTIES_LOCATION = {
714            **generator.Generator.PROPERTIES_LOCATION,
715            exp.PartitionedByProperty: exp.Properties.Location.POST_SCHEMA,
716            exp.VolatileProperty: exp.Properties.Location.UNSUPPORTED,
717        }
718
719        # from: https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#reserved_keywords
720        RESERVED_KEYWORDS = {
721            "all",
722            "and",
723            "any",
724            "array",
725            "as",
726            "asc",
727            "assert_rows_modified",
728            "at",
729            "between",
730            "by",
731            "case",
732            "cast",
733            "collate",
734            "contains",
735            "create",
736            "cross",
737            "cube",
738            "current",
739            "default",
740            "define",
741            "desc",
742            "distinct",
743            "else",
744            "end",
745            "enum",
746            "escape",
747            "except",
748            "exclude",
749            "exists",
750            "extract",
751            "false",
752            "fetch",
753            "following",
754            "for",
755            "from",
756            "full",
757            "group",
758            "grouping",
759            "groups",
760            "hash",
761            "having",
762            "if",
763            "ignore",
764            "in",
765            "inner",
766            "intersect",
767            "interval",
768            "into",
769            "is",
770            "join",
771            "lateral",
772            "left",
773            "like",
774            "limit",
775            "lookup",
776            "merge",
777            "natural",
778            "new",
779            "no",
780            "not",
781            "null",
782            "nulls",
783            "of",
784            "on",
785            "or",
786            "order",
787            "outer",
788            "over",
789            "partition",
790            "preceding",
791            "proto",
792            "qualify",
793            "range",
794            "recursive",
795            "respect",
796            "right",
797            "rollup",
798            "rows",
799            "select",
800            "set",
801            "some",
802            "struct",
803            "tablesample",
804            "then",
805            "to",
806            "treat",
807            "true",
808            "unbounded",
809            "union",
810            "unnest",
811            "using",
812            "when",
813            "where",
814            "window",
815            "with",
816            "within",
817        }
818
819        def mod_sql(self, expression: exp.Mod) -> str:
820            this = expression.this
821            expr = expression.expression
822            return self.func(
823                "MOD",
824                this.unnest() if isinstance(this, exp.Paren) else this,
825                expr.unnest() if isinstance(expr, exp.Paren) else expr,
826            )
827
828        def column_parts(self, expression: exp.Column) -> str:
829            if expression.meta.get("quoted_column"):
830                # If a column reference is of the form `dataset.table`.name, we need
831                # to preserve the quoted table path, otherwise the reference breaks
832                table_parts = ".".join(p.name for p in expression.parts[:-1])
833                table_path = self.sql(exp.Identifier(this=table_parts, quoted=True))
834                return f"{table_path}.{self.sql(expression, 'this')}"
835
836            return super().column_parts(expression)
837
838        def table_parts(self, expression: exp.Table) -> str:
839            # Depending on the context, `x.y` may not resolve to the same data source as `x`.`y`, so
840            # we need to make sure the correct quoting is used in each case.
841            #
842            # For example, if there is a CTE x that clashes with a schema name, then the former will
843            # return the table y in that schema, whereas the latter will return the CTE's y column:
844            #
845            # - WITH x AS (SELECT [1, 2] AS y) SELECT * FROM x, `x.y`   -> cross join
846            # - WITH x AS (SELECT [1, 2] AS y) SELECT * FROM x, `x`.`y` -> implicit unnest
847            if expression.meta.get("quoted_table"):
848                table_parts = ".".join(p.name for p in expression.parts)
849                return self.sql(exp.Identifier(this=table_parts, quoted=True))
850
851            return super().table_parts(expression)
852
853        def timetostr_sql(self, expression: exp.TimeToStr) -> str:
854            this = expression.this if isinstance(expression.this, exp.TsOrDsToDate) else expression
855            return self.func("FORMAT_DATE", self.format_time(expression), this.this)
856
857        def eq_sql(self, expression: exp.EQ) -> str:
858            # Operands of = cannot be NULL in BigQuery
859            if isinstance(expression.left, exp.Null) or isinstance(expression.right, exp.Null):
860                if not isinstance(expression.parent, exp.Update):
861                    return "NULL"
862
863            return self.binary(expression, "=")
864
865        def attimezone_sql(self, expression: exp.AtTimeZone) -> str:
866            parent = expression.parent
867
868            # BigQuery allows CAST(.. AS {STRING|TIMESTAMP} [FORMAT <fmt> [AT TIME ZONE <tz>]]).
869            # Only the TIMESTAMP one should use the below conversion, when AT TIME ZONE is included.
870            if not isinstance(parent, exp.Cast) or not parent.to.is_type("text"):
871                return self.func(
872                    "TIMESTAMP", self.func("DATETIME", expression.this, expression.args.get("zone"))
873                )
874
875            return super().attimezone_sql(expression)
876
877        def trycast_sql(self, expression: exp.TryCast) -> str:
878            return self.cast_sql(expression, safe_prefix="SAFE_")
879
880        def bracket_sql(self, expression: exp.Bracket) -> str:
881            this = expression.this
882            expressions = expression.expressions
883
884            if len(expressions) == 1 and this and this.is_type(exp.DataType.Type.STRUCT):
885                arg = expressions[0]
886                if arg.type is None:
887                    from sqlglot.optimizer.annotate_types import annotate_types
888
889                    arg = annotate_types(arg)
890
891                if arg.type and arg.type.this in exp.DataType.TEXT_TYPES:
892                    # BQ doesn't support bracket syntax with string values for structs
893                    return f"{self.sql(this)}.{arg.name}"
894
895            expressions_sql = self.expressions(expression, flat=True)
896            offset = expression.args.get("offset")
897
898            if offset == 0:
899                expressions_sql = f"OFFSET({expressions_sql})"
900            elif offset == 1:
901                expressions_sql = f"ORDINAL({expressions_sql})"
902            elif offset is not None:
903                self.unsupported(f"Unsupported array offset: {offset}")
904
905            if expression.args.get("safe"):
906                expressions_sql = f"SAFE_{expressions_sql}"
907
908            return f"{self.sql(this)}[{expressions_sql}]"
909
910        def in_unnest_op(self, expression: exp.Unnest) -> str:
911            return self.sql(expression)
912
913        def except_op(self, expression: exp.Except) -> str:
914            if not expression.args.get("distinct"):
915                self.unsupported("EXCEPT without DISTINCT is not supported in BigQuery")
916            return f"EXCEPT{' DISTINCT' if expression.args.get('distinct') else ' ALL'}"
917
918        def intersect_op(self, expression: exp.Intersect) -> str:
919            if not expression.args.get("distinct"):
920                self.unsupported("INTERSECT without DISTINCT is not supported in BigQuery")
921            return f"INTERSECT{' DISTINCT' if expression.args.get('distinct') else ' ALL'}"
922
923        def version_sql(self, expression: exp.Version) -> str:
924            if expression.name == "TIMESTAMP":
925                expression.set("this", "SYSTEM_TIME")
926            return super().version_sql(expression)
WEEK_OFFSET = -1

First day of the week in DATE_TRUNC(week). Defaults to 0 (Monday). -1 would be Sunday.

UNNEST_COLUMN_ONLY = True

Whether UNNEST table aliases are treated as column aliases.

SUPPORTS_USER_DEFINED_TYPES = False

Whether user-defined data types are supported.

SUPPORTS_SEMI_ANTI_JOIN = False

Whether SEMI or ANTI joins are supported.

LOG_BASE_FIRST: Optional[bool] = False

Whether the base comes first in the LOG function. Possible values: True, False, None (two arguments are not supported by LOG)

HEX_LOWERCASE = True

Whether the HEX function returns a lowercase hexadecimal string.

NORMALIZATION_STRATEGY = <NormalizationStrategy.CASE_INSENSITIVE: 'CASE_INSENSITIVE'>

Specifies the strategy according to which identifiers should be normalized.

NORMALIZE_FUNCTIONS: bool | str = False

Determines how function names are going to be normalized.

Possible values:

"upper" or True: Convert names to uppercase. "lower": Convert names to lowercase. False: Disables function name normalization.

TIME_MAPPING: Dict[str, str] = {'%D': '%m/%d/%y', '%E6S': '%S.%f'}

Associates this dialect's time formats with their equivalent Python strftime formats.

FORMAT_MAPPING: Dict[str, str] = {'DD': '%d', 'MM': '%m', 'MON': '%b', 'MONTH': '%B', 'YYYY': '%Y', 'YY': '%y', 'HH': '%I', 'HH12': '%I', 'HH24': '%H', 'MI': '%M', 'SS': '%S', 'SSSSS': '%f', 'TZH': '%z'}

Helper which is used for parsing the special syntax CAST(x AS DATE FORMAT 'yyyy'). If empty, the corresponding trie will be constructed off of TIME_MAPPING.

PSEUDOCOLUMNS: Set[str] = {'_PARTITIONTIME', '_PARTITIONDATE'}

Columns that are auto-generated by the engine corresponding to this dialect. For example, such columns may be excluded from SELECT * queries.

def normalize_identifier(self, expression: ~E) -> ~E:
258    def normalize_identifier(self, expression: E) -> E:
259        if (
260            isinstance(expression, exp.Identifier)
261            and self.normalization_strategy is not NormalizationStrategy.CASE_SENSITIVE
262        ):
263            parent = expression.parent
264            while isinstance(parent, exp.Dot):
265                parent = parent.parent
266
267            # In BigQuery, CTEs are case-insensitive, but UDF and table names are case-sensitive
268            # by default. The following check uses a heuristic to detect tables based on whether
269            # they are qualified. This should generally be correct, because tables in BigQuery
270            # must be qualified with at least a dataset, unless @@dataset_id is set.
271            case_sensitive = (
272                isinstance(parent, exp.UserDefinedFunction)
273                or (
274                    isinstance(parent, exp.Table)
275                    and parent.db
276                    and (parent.meta.get("quoted_table") or not parent.meta.get("maybe_column"))
277                )
278                or expression.meta.get("is_table")
279            )
280            if not case_sensitive:
281                expression.set("this", expression.this.lower())
282
283        return expression

Transforms an identifier in a way that resembles how it'd be resolved by this dialect.

For example, an identifier like FoO would be resolved as foo in Postgres, because it lowercases all unquoted identifiers. On the other hand, Snowflake uppercases them, so it would resolve it as FOO. If it was quoted, it'd need to be treated as case-sensitive, and so any normalization would be prohibited in order to avoid "breaking" the identifier.

There are also dialects like Spark, which are case-insensitive even when quotes are present, and dialects like MySQL, whose resolution rules match those employed by the underlying operating system, for example they may always be case-sensitive in Linux.

Finally, the normalization behavior of some engines can even be controlled through flags, like in Redshift's case, where users can explicitly set enable_case_sensitive_identifier.

SQLGlot aims to understand and handle all of these different behaviors gracefully, so that it can analyze queries in the optimizer and successfully capture their semantics.

SUPPORTS_COLUMN_JOIN_MARKS = False

Whether the old-style outer join (+) syntax is supported.

UNESCAPED_SEQUENCES: Dict[str, str] = {'\\a': '\x07', '\\b': '\x08', '\\f': '\x0c', '\\n': '\n', '\\r': '\r', '\\t': '\t', '\\v': '\x0b', '\\\\': '\\'}

Mapping of an escaped sequence (\n) to its unescaped version ( ).

tokenizer_class = <class 'BigQuery.Tokenizer'>
jsonpath_tokenizer_class = <class 'sqlglot.tokens.JSONPathTokenizer'>
parser_class = <class 'BigQuery.Parser'>
generator_class = <class 'BigQuery.Generator'>
TIME_TRIE: Dict = {'%': {'D': {0: True}, 'E': {'6': {'S': {0: True}}}}}
FORMAT_TRIE: Dict = {'D': {'D': {0: True}}, 'M': {'M': {0: True}, 'O': {'N': {0: True, 'T': {'H': {0: True}}}}, 'I': {0: True}}, 'Y': {'Y': {'Y': {'Y': {0: True}}, 0: True}}, 'H': {'H': {0: True, '1': {'2': {0: True}}, '2': {'4': {0: True}}}}, 'S': {'S': {0: True, 'S': {'S': {'S': {0: True}}}}}, 'T': {'Z': {'H': {0: True}}}}
INVERSE_TIME_MAPPING: Dict[str, str] = {'%m/%d/%y': '%D', '%S.%f': '%E6S'}
INVERSE_TIME_TRIE: Dict = {'%': {'m': {'/': {'%': {'d': {'/': {'%': {'y': {0: True}}}}}}}, 'S': {'.': {'%': {'f': {0: True}}}}}}
ESCAPED_SEQUENCES: Dict[str, str] = {'\x07': '\\a', '\x08': '\\b', '\x0c': '\\f', '\n': '\\n', '\r': '\\r', '\t': '\\t', '\x0b': '\\v', '\\': '\\\\'}
QUOTE_START = "'"
QUOTE_END = "'"
IDENTIFIER_START = '`'
IDENTIFIER_END = '`'
BIT_START: Optional[str] = None
BIT_END: Optional[str] = None
HEX_START: Optional[str] = '0x'
HEX_END: Optional[str] = ''
BYTE_START: Optional[str] = "b'"
BYTE_END: Optional[str] = "'"
UNICODE_START: Optional[str] = None
UNICODE_END: Optional[str] = None
class BigQuery.Tokenizer(sqlglot.tokens.Tokenizer):
285    class Tokenizer(tokens.Tokenizer):
286        QUOTES = ["'", '"', '"""', "'''"]
287        COMMENTS = ["--", "#", ("/*", "*/")]
288        IDENTIFIERS = ["`"]
289        STRING_ESCAPES = ["\\"]
290
291        HEX_STRINGS = [("0x", ""), ("0X", "")]
292
293        BYTE_STRINGS = [
294            (prefix + q, q) for q in t.cast(t.List[str], QUOTES) for prefix in ("b", "B")
295        ]
296
297        RAW_STRINGS = [
298            (prefix + q, q) for q in t.cast(t.List[str], QUOTES) for prefix in ("r", "R")
299        ]
300
301        KEYWORDS = {
302            **tokens.Tokenizer.KEYWORDS,
303            "ANY TYPE": TokenType.VARIANT,
304            "BEGIN": TokenType.COMMAND,
305            "BEGIN TRANSACTION": TokenType.BEGIN,
306            "BYTES": TokenType.BINARY,
307            "CURRENT_DATETIME": TokenType.CURRENT_DATETIME,
308            "DATETIME": TokenType.TIMESTAMP,
309            "DECLARE": TokenType.COMMAND,
310            "ELSEIF": TokenType.COMMAND,
311            "EXCEPTION": TokenType.COMMAND,
312            "FLOAT64": TokenType.DOUBLE,
313            "FOR SYSTEM_TIME": TokenType.TIMESTAMP_SNAPSHOT,
314            "MODEL": TokenType.MODEL,
315            "NOT DETERMINISTIC": TokenType.VOLATILE,
316            "RECORD": TokenType.STRUCT,
317            "TIMESTAMP": TokenType.TIMESTAMPTZ,
318        }
319        KEYWORDS.pop("DIV")
320        KEYWORDS.pop("VALUES")
QUOTES = ["'", '"', '"""', "'''"]
COMMENTS = ['--', '#', ('/*', '*/')]
IDENTIFIERS = ['`']
STRING_ESCAPES = ['\\']
HEX_STRINGS = [('0x', ''), ('0X', '')]
BYTE_STRINGS = [("b'", "'"), ("B'", "'"), ('b"', '"'), ('B"', '"'), ('b"""', '"""'), ('B"""', '"""'), ("b'''", "'''"), ("B'''", "'''")]
RAW_STRINGS = [("r'", "'"), ("R'", "'"), ('r"', '"'), ('R"', '"'), ('r"""', '"""'), ('R"""', '"""'), ("r'''", "'''"), ("R'''", "'''")]
KEYWORDS = {'{%': <TokenType.BLOCK_START: 'BLOCK_START'>, '{%+': <TokenType.BLOCK_START: 'BLOCK_START'>, '{%-': <TokenType.BLOCK_START: 'BLOCK_START'>, '%}': <TokenType.BLOCK_END: 'BLOCK_END'>, '+%}': <TokenType.BLOCK_END: 'BLOCK_END'>, '-%}': <TokenType.BLOCK_END: 'BLOCK_END'>, '{{+': <TokenType.BLOCK_START: 'BLOCK_START'>, '{{-': <TokenType.BLOCK_START: 'BLOCK_START'>, '+}}': <TokenType.BLOCK_END: 'BLOCK_END'>, '-}}': <TokenType.BLOCK_END: 'BLOCK_END'>, '/*+': <TokenType.HINT: 'HINT'>, '==': <TokenType.EQ: 'EQ'>, '::': <TokenType.DCOLON: 'DCOLON'>, '||': <TokenType.DPIPE: 'DPIPE'>, '>=': <TokenType.GTE: 'GTE'>, '<=': <TokenType.LTE: 'LTE'>, '<>': <TokenType.NEQ: 'NEQ'>, '!=': <TokenType.NEQ: 'NEQ'>, ':=': <TokenType.COLON_EQ: 'COLON_EQ'>, '<=>': <TokenType.NULLSAFE_EQ: 'NULLSAFE_EQ'>, '->': <TokenType.ARROW: 'ARROW'>, '->>': <TokenType.DARROW: 'DARROW'>, '=>': <TokenType.FARROW: 'FARROW'>, '#>': <TokenType.HASH_ARROW: 'HASH_ARROW'>, '#>>': <TokenType.DHASH_ARROW: 'DHASH_ARROW'>, '<->': <TokenType.LR_ARROW: 'LR_ARROW'>, '&&': <TokenType.DAMP: 'DAMP'>, '??': <TokenType.DQMARK: 'DQMARK'>, 'ALL': <TokenType.ALL: 'ALL'>, 'ALWAYS': <TokenType.ALWAYS: 'ALWAYS'>, 'AND': <TokenType.AND: 'AND'>, 'ANTI': <TokenType.ANTI: 'ANTI'>, 'ANY': <TokenType.ANY: 'ANY'>, 'ASC': <TokenType.ASC: 'ASC'>, 'AS': <TokenType.ALIAS: 'ALIAS'>, 'ASOF': <TokenType.ASOF: 'ASOF'>, 'AUTOINCREMENT': <TokenType.AUTO_INCREMENT: 'AUTO_INCREMENT'>, 'AUTO_INCREMENT': <TokenType.AUTO_INCREMENT: 'AUTO_INCREMENT'>, 'BEGIN': <TokenType.COMMAND: 'COMMAND'>, 'BETWEEN': <TokenType.BETWEEN: 'BETWEEN'>, 'CACHE': <TokenType.CACHE: 'CACHE'>, 'UNCACHE': <TokenType.UNCACHE: 'UNCACHE'>, 'CASE': <TokenType.CASE: 'CASE'>, 'CHARACTER SET': <TokenType.CHARACTER_SET: 'CHARACTER_SET'>, 'CLUSTER BY': <TokenType.CLUSTER_BY: 'CLUSTER_BY'>, 'COLLATE': <TokenType.COLLATE: 'COLLATE'>, 'COLUMN': <TokenType.COLUMN: 'COLUMN'>, 'COMMIT': <TokenType.COMMIT: 'COMMIT'>, 'CONNECT BY': <TokenType.CONNECT_BY: 'CONNECT_BY'>, 'CONSTRAINT': <TokenType.CONSTRAINT: 'CONSTRAINT'>, 'COPY': <TokenType.COPY: 'COPY'>, 'CREATE': <TokenType.CREATE: 'CREATE'>, 'CROSS': <TokenType.CROSS: 'CROSS'>, 'CUBE': <TokenType.CUBE: 'CUBE'>, 'CURRENT_DATE': <TokenType.CURRENT_DATE: 'CURRENT_DATE'>, 'CURRENT_TIME': <TokenType.CURRENT_TIME: 'CURRENT_TIME'>, 'CURRENT_TIMESTAMP': <TokenType.CURRENT_TIMESTAMP: 'CURRENT_TIMESTAMP'>, 'CURRENT_USER': <TokenType.CURRENT_USER: 'CURRENT_USER'>, 'DATABASE': <TokenType.DATABASE: 'DATABASE'>, 'DEFAULT': <TokenType.DEFAULT: 'DEFAULT'>, 'DELETE': <TokenType.DELETE: 'DELETE'>, 'DESC': <TokenType.DESC: 'DESC'>, 'DESCRIBE': <TokenType.DESCRIBE: 'DESCRIBE'>, 'DISTINCT': <TokenType.DISTINCT: 'DISTINCT'>, 'DISTRIBUTE BY': <TokenType.DISTRIBUTE_BY: 'DISTRIBUTE_BY'>, 'DROP': <TokenType.DROP: 'DROP'>, 'ELSE': <TokenType.ELSE: 'ELSE'>, 'END': <TokenType.END: 'END'>, 'ENUM': <TokenType.ENUM: 'ENUM'>, 'ESCAPE': <TokenType.ESCAPE: 'ESCAPE'>, 'EXCEPT': <TokenType.EXCEPT: 'EXCEPT'>, 'EXECUTE': <TokenType.EXECUTE: 'EXECUTE'>, 'EXISTS': <TokenType.EXISTS: 'EXISTS'>, 'FALSE': <TokenType.FALSE: 'FALSE'>, 'FETCH': <TokenType.FETCH: 'FETCH'>, 'FILTER': <TokenType.FILTER: 'FILTER'>, 'FIRST': <TokenType.FIRST: 'FIRST'>, 'FULL': <TokenType.FULL: 'FULL'>, 'FUNCTION': <TokenType.FUNCTION: 'FUNCTION'>, 'FOR': <TokenType.FOR: 'FOR'>, 'FOREIGN KEY': <TokenType.FOREIGN_KEY: 'FOREIGN_KEY'>, 'FORMAT': <TokenType.FORMAT: 'FORMAT'>, 'FROM': <TokenType.FROM: 'FROM'>, 'GEOGRAPHY': <TokenType.GEOGRAPHY: 'GEOGRAPHY'>, 'GEOMETRY': <TokenType.GEOMETRY: 'GEOMETRY'>, 'GLOB': <TokenType.GLOB: 'GLOB'>, 'GROUP BY': <TokenType.GROUP_BY: 'GROUP_BY'>, 'GROUPING SETS': <TokenType.GROUPING_SETS: 'GROUPING_SETS'>, 'HAVING': <TokenType.HAVING: 'HAVING'>, 'ILIKE': <TokenType.ILIKE: 'ILIKE'>, 'IN': <TokenType.IN: 'IN'>, 'INDEX': <TokenType.INDEX: 'INDEX'>, 'INET': <TokenType.INET: 'INET'>, 'INNER': <TokenType.INNER: 'INNER'>, 'INSERT': <TokenType.INSERT: 'INSERT'>, 'INTERVAL': <TokenType.INTERVAL: 'INTERVAL'>, 'INTERSECT': <TokenType.INTERSECT: 'INTERSECT'>, 'INTO': <TokenType.INTO: 'INTO'>, 'IS': <TokenType.IS: 'IS'>, 'ISNULL': <TokenType.ISNULL: 'ISNULL'>, 'JOIN': <TokenType.JOIN: 'JOIN'>, 'KEEP': <TokenType.KEEP: 'KEEP'>, 'KILL': <TokenType.KILL: 'KILL'>, 'LATERAL': <TokenType.LATERAL: 'LATERAL'>, 'LEFT': <TokenType.LEFT: 'LEFT'>, 'LIKE': <TokenType.LIKE: 'LIKE'>, 'LIMIT': <TokenType.LIMIT: 'LIMIT'>, 'LOAD': <TokenType.LOAD: 'LOAD'>, 'LOCK': <TokenType.LOCK: 'LOCK'>, 'MERGE': <TokenType.MERGE: 'MERGE'>, 'NATURAL': <TokenType.NATURAL: 'NATURAL'>, 'NEXT': <TokenType.NEXT: 'NEXT'>, 'NOT': <TokenType.NOT: 'NOT'>, 'NOTNULL': <TokenType.NOTNULL: 'NOTNULL'>, 'NULL': <TokenType.NULL: 'NULL'>, 'OBJECT': <TokenType.OBJECT: 'OBJECT'>, 'OFFSET': <TokenType.OFFSET: 'OFFSET'>, 'ON': <TokenType.ON: 'ON'>, 'OR': <TokenType.OR: 'OR'>, 'XOR': <TokenType.XOR: 'XOR'>, 'ORDER BY': <TokenType.ORDER_BY: 'ORDER_BY'>, 'ORDINALITY': <TokenType.ORDINALITY: 'ORDINALITY'>, 'OUTER': <TokenType.OUTER: 'OUTER'>, 'OVER': <TokenType.OVER: 'OVER'>, 'OVERLAPS': <TokenType.OVERLAPS: 'OVERLAPS'>, 'OVERWRITE': <TokenType.OVERWRITE: 'OVERWRITE'>, 'PARTITION': <TokenType.PARTITION: 'PARTITION'>, 'PARTITION BY': <TokenType.PARTITION_BY: 'PARTITION_BY'>, 'PARTITIONED BY': <TokenType.PARTITION_BY: 'PARTITION_BY'>, 'PARTITIONED_BY': <TokenType.PARTITION_BY: 'PARTITION_BY'>, 'PERCENT': <TokenType.PERCENT: 'PERCENT'>, 'PIVOT': <TokenType.PIVOT: 'PIVOT'>, 'PRAGMA': <TokenType.PRAGMA: 'PRAGMA'>, 'PRIMARY KEY': <TokenType.PRIMARY_KEY: 'PRIMARY_KEY'>, 'PROCEDURE': <TokenType.PROCEDURE: 'PROCEDURE'>, 'QUALIFY': <TokenType.QUALIFY: 'QUALIFY'>, 'RANGE': <TokenType.RANGE: 'RANGE'>, 'RECURSIVE': <TokenType.RECURSIVE: 'RECURSIVE'>, 'REGEXP': <TokenType.RLIKE: 'RLIKE'>, 'REPLACE': <TokenType.REPLACE: 'REPLACE'>, 'RETURNING': <TokenType.RETURNING: 'RETURNING'>, 'REFERENCES': <TokenType.REFERENCES: 'REFERENCES'>, 'RIGHT': <TokenType.RIGHT: 'RIGHT'>, 'RLIKE': <TokenType.RLIKE: 'RLIKE'>, 'ROLLBACK': <TokenType.ROLLBACK: 'ROLLBACK'>, 'ROLLUP': <TokenType.ROLLUP: 'ROLLUP'>, 'ROW': <TokenType.ROW: 'ROW'>, 'ROWS': <TokenType.ROWS: 'ROWS'>, 'SCHEMA': <TokenType.SCHEMA: 'SCHEMA'>, 'SELECT': <TokenType.SELECT: 'SELECT'>, 'SEMI': <TokenType.SEMI: 'SEMI'>, 'SET': <TokenType.SET: 'SET'>, 'SETTINGS': <TokenType.SETTINGS: 'SETTINGS'>, 'SHOW': <TokenType.SHOW: 'SHOW'>, 'SIMILAR TO': <TokenType.SIMILAR_TO: 'SIMILAR_TO'>, 'SOME': <TokenType.SOME: 'SOME'>, 'SORT BY': <TokenType.SORT_BY: 'SORT_BY'>, 'START WITH': <TokenType.START_WITH: 'START_WITH'>, 'STRAIGHT_JOIN': <TokenType.STRAIGHT_JOIN: 'STRAIGHT_JOIN'>, 'TABLE': <TokenType.TABLE: 'TABLE'>, 'TABLESAMPLE': <TokenType.TABLE_SAMPLE: 'TABLE_SAMPLE'>, 'TEMP': <TokenType.TEMPORARY: 'TEMPORARY'>, 'TEMPORARY': <TokenType.TEMPORARY: 'TEMPORARY'>, 'THEN': <TokenType.THEN: 'THEN'>, 'TRUE': <TokenType.TRUE: 'TRUE'>, 'TRUNCATE': <TokenType.TRUNCATE: 'TRUNCATE'>, 'UNION': <TokenType.UNION: 'UNION'>, 'UNKNOWN': <TokenType.UNKNOWN: 'UNKNOWN'>, 'UNNEST': <TokenType.UNNEST: 'UNNEST'>, 'UNPIVOT': <TokenType.UNPIVOT: 'UNPIVOT'>, 'UPDATE': <TokenType.UPDATE: 'UPDATE'>, 'USE': <TokenType.USE: 'USE'>, 'USING': <TokenType.USING: 'USING'>, 'UUID': <TokenType.UUID: 'UUID'>, 'VIEW': <TokenType.VIEW: 'VIEW'>, 'VOLATILE': <TokenType.VOLATILE: 'VOLATILE'>, 'WHEN': <TokenType.WHEN: 'WHEN'>, 'WHERE': <TokenType.WHERE: 'WHERE'>, 'WINDOW': <TokenType.WINDOW: 'WINDOW'>, 'WITH': <TokenType.WITH: 'WITH'>, 'APPLY': <TokenType.APPLY: 'APPLY'>, 'ARRAY': <TokenType.ARRAY: 'ARRAY'>, 'BIT': <TokenType.BIT: 'BIT'>, 'BOOL': <TokenType.BOOLEAN: 'BOOLEAN'>, 'BOOLEAN': <TokenType.BOOLEAN: 'BOOLEAN'>, 'BYTE': <TokenType.TINYINT: 'TINYINT'>, 'MEDIUMINT': <TokenType.MEDIUMINT: 'MEDIUMINT'>, 'INT1': <TokenType.TINYINT: 'TINYINT'>, 'TINYINT': <TokenType.TINYINT: 'TINYINT'>, 'INT16': <TokenType.SMALLINT: 'SMALLINT'>, 'SHORT': <TokenType.SMALLINT: 'SMALLINT'>, 'SMALLINT': <TokenType.SMALLINT: 'SMALLINT'>, 'INT128': <TokenType.INT128: 'INT128'>, 'HUGEINT': <TokenType.INT128: 'INT128'>, 'INT2': <TokenType.SMALLINT: 'SMALLINT'>, 'INTEGER': <TokenType.INT: 'INT'>, 'INT': <TokenType.INT: 'INT'>, 'INT4': <TokenType.INT: 'INT'>, 'INT32': <TokenType.INT: 'INT'>, 'INT64': <TokenType.BIGINT: 'BIGINT'>, 'LONG': <TokenType.BIGINT: 'BIGINT'>, 'BIGINT': <TokenType.BIGINT: 'BIGINT'>, 'INT8': <TokenType.TINYINT: 'TINYINT'>, 'UINT': <TokenType.UINT: 'UINT'>, 'DEC': <TokenType.DECIMAL: 'DECIMAL'>, 'DECIMAL': <TokenType.DECIMAL: 'DECIMAL'>, 'BIGDECIMAL': <TokenType.BIGDECIMAL: 'BIGDECIMAL'>, 'BIGNUMERIC': <TokenType.BIGDECIMAL: 'BIGDECIMAL'>, 'LIST': <TokenType.LIST: 'LIST'>, 'MAP': <TokenType.MAP: 'MAP'>, 'NULLABLE': <TokenType.NULLABLE: 'NULLABLE'>, 'NUMBER': <TokenType.DECIMAL: 'DECIMAL'>, 'NUMERIC': <TokenType.DECIMAL: 'DECIMAL'>, 'FIXED': <TokenType.DECIMAL: 'DECIMAL'>, 'REAL': <TokenType.FLOAT: 'FLOAT'>, 'FLOAT': <TokenType.FLOAT: 'FLOAT'>, 'FLOAT4': <TokenType.FLOAT: 'FLOAT'>, 'FLOAT8': <TokenType.DOUBLE: 'DOUBLE'>, 'DOUBLE': <TokenType.DOUBLE: 'DOUBLE'>, 'DOUBLE PRECISION': <TokenType.DOUBLE: 'DOUBLE'>, 'JSON': <TokenType.JSON: 'JSON'>, 'JSONB': <TokenType.JSONB: 'JSONB'>, 'CHAR': <TokenType.CHAR: 'CHAR'>, 'CHARACTER': <TokenType.CHAR: 'CHAR'>, 'NCHAR': <TokenType.NCHAR: 'NCHAR'>, 'VARCHAR': <TokenType.VARCHAR: 'VARCHAR'>, 'VARCHAR2': <TokenType.VARCHAR: 'VARCHAR'>, 'NVARCHAR': <TokenType.NVARCHAR: 'NVARCHAR'>, 'NVARCHAR2': <TokenType.NVARCHAR: 'NVARCHAR'>, 'BPCHAR': <TokenType.BPCHAR: 'BPCHAR'>, 'STR': <TokenType.TEXT: 'TEXT'>, 'STRING': <TokenType.TEXT: 'TEXT'>, 'TEXT': <TokenType.TEXT: 'TEXT'>, 'LONGTEXT': <TokenType.LONGTEXT: 'LONGTEXT'>, 'MEDIUMTEXT': <TokenType.MEDIUMTEXT: 'MEDIUMTEXT'>, 'TINYTEXT': <TokenType.TINYTEXT: 'TINYTEXT'>, 'CLOB': <TokenType.TEXT: 'TEXT'>, 'LONGVARCHAR': <TokenType.TEXT: 'TEXT'>, 'BINARY': <TokenType.BINARY: 'BINARY'>, 'BLOB': <TokenType.VARBINARY: 'VARBINARY'>, 'LONGBLOB': <TokenType.LONGBLOB: 'LONGBLOB'>, 'MEDIUMBLOB': <TokenType.MEDIUMBLOB: 'MEDIUMBLOB'>, 'TINYBLOB': <TokenType.TINYBLOB: 'TINYBLOB'>, 'BYTEA': <TokenType.VARBINARY: 'VARBINARY'>, 'VARBINARY': <TokenType.VARBINARY: 'VARBINARY'>, 'TIME': <TokenType.TIME: 'TIME'>, 'TIMETZ': <TokenType.TIMETZ: 'TIMETZ'>, 'TIMESTAMP': <TokenType.TIMESTAMPTZ: 'TIMESTAMPTZ'>, 'TIMESTAMPTZ': <TokenType.TIMESTAMPTZ: 'TIMESTAMPTZ'>, 'TIMESTAMPLTZ': <TokenType.TIMESTAMPLTZ: 'TIMESTAMPLTZ'>, 'TIMESTAMP_LTZ': <TokenType.TIMESTAMPLTZ: 'TIMESTAMPLTZ'>, 'TIMESTAMPNTZ': <TokenType.TIMESTAMPNTZ: 'TIMESTAMPNTZ'>, 'TIMESTAMP_NTZ': <TokenType.TIMESTAMPNTZ: 'TIMESTAMPNTZ'>, 'DATE': <TokenType.DATE: 'DATE'>, 'DATETIME': <TokenType.TIMESTAMP: 'TIMESTAMP'>, 'INT4RANGE': <TokenType.INT4RANGE: 'INT4RANGE'>, 'INT4MULTIRANGE': <TokenType.INT4MULTIRANGE: 'INT4MULTIRANGE'>, 'INT8RANGE': <TokenType.INT8RANGE: 'INT8RANGE'>, 'INT8MULTIRANGE': <TokenType.INT8MULTIRANGE: 'INT8MULTIRANGE'>, 'NUMRANGE': <TokenType.NUMRANGE: 'NUMRANGE'>, 'NUMMULTIRANGE': <TokenType.NUMMULTIRANGE: 'NUMMULTIRANGE'>, 'TSRANGE': <TokenType.TSRANGE: 'TSRANGE'>, 'TSMULTIRANGE': <TokenType.TSMULTIRANGE: 'TSMULTIRANGE'>, 'TSTZRANGE': <TokenType.TSTZRANGE: 'TSTZRANGE'>, 'TSTZMULTIRANGE': <TokenType.TSTZMULTIRANGE: 'TSTZMULTIRANGE'>, 'DATERANGE': <TokenType.DATERANGE: 'DATERANGE'>, 'DATEMULTIRANGE': <TokenType.DATEMULTIRANGE: 'DATEMULTIRANGE'>, 'UNIQUE': <TokenType.UNIQUE: 'UNIQUE'>, 'STRUCT': <TokenType.STRUCT: 'STRUCT'>, 'SEQUENCE': <TokenType.SEQUENCE: 'SEQUENCE'>, 'VARIANT': <TokenType.VARIANT: 'VARIANT'>, 'ALTER': <TokenType.ALTER: 'ALTER'>, 'ANALYZE': <TokenType.COMMAND: 'COMMAND'>, 'CALL': <TokenType.COMMAND: 'COMMAND'>, 'COMMENT': <TokenType.COMMENT: 'COMMENT'>, 'EXPLAIN': <TokenType.COMMAND: 'COMMAND'>, 'GRANT': <TokenType.COMMAND: 'COMMAND'>, 'OPTIMIZE': <TokenType.COMMAND: 'COMMAND'>, 'PREPARE': <TokenType.COMMAND: 'COMMAND'>, 'VACUUM': <TokenType.COMMAND: 'COMMAND'>, 'USER-DEFINED': <TokenType.USERDEFINED: 'USERDEFINED'>, 'FOR VERSION': <TokenType.VERSION_SNAPSHOT: 'VERSION_SNAPSHOT'>, 'FOR TIMESTAMP': <TokenType.TIMESTAMP_SNAPSHOT: 'TIMESTAMP_SNAPSHOT'>, 'ANY TYPE': <TokenType.VARIANT: 'VARIANT'>, 'BEGIN TRANSACTION': <TokenType.BEGIN: 'BEGIN'>, 'BYTES': <TokenType.BINARY: 'BINARY'>, 'CURRENT_DATETIME': <TokenType.CURRENT_DATETIME: 'CURRENT_DATETIME'>, 'DECLARE': <TokenType.COMMAND: 'COMMAND'>, 'ELSEIF': <TokenType.COMMAND: 'COMMAND'>, 'EXCEPTION': <TokenType.COMMAND: 'COMMAND'>, 'FLOAT64': <TokenType.DOUBLE: 'DOUBLE'>, 'FOR SYSTEM_TIME': <TokenType.TIMESTAMP_SNAPSHOT: 'TIMESTAMP_SNAPSHOT'>, 'MODEL': <TokenType.MODEL: 'MODEL'>, 'NOT DETERMINISTIC': <TokenType.VOLATILE: 'VOLATILE'>, 'RECORD': <TokenType.STRUCT: 'STRUCT'>}
class BigQuery.Parser(sqlglot.parser.Parser):
322    class Parser(parser.Parser):
323        PREFIXED_PIVOT_COLUMNS = True
324        LOG_DEFAULTS_TO_LN = True
325        SUPPORTS_IMPLICIT_UNNEST = True
326
327        FUNCTIONS = {
328            **parser.Parser.FUNCTIONS,
329            "DATE": _build_date,
330            "DATE_ADD": build_date_delta_with_interval(exp.DateAdd),
331            "DATE_SUB": build_date_delta_with_interval(exp.DateSub),
332            "DATE_TRUNC": lambda args: exp.DateTrunc(
333                unit=exp.Literal.string(str(seq_get(args, 1))),
334                this=seq_get(args, 0),
335            ),
336            "DATETIME": _build_datetime,
337            "DATETIME_ADD": build_date_delta_with_interval(exp.DatetimeAdd),
338            "DATETIME_SUB": build_date_delta_with_interval(exp.DatetimeSub),
339            "DIV": binary_from_function(exp.IntDiv),
340            "FORMAT_DATE": lambda args: exp.TimeToStr(
341                this=exp.TsOrDsToDate(this=seq_get(args, 1)), format=seq_get(args, 0)
342            ),
343            "GENERATE_ARRAY": exp.GenerateSeries.from_arg_list,
344            "JSON_EXTRACT_SCALAR": lambda args: exp.JSONExtractScalar(
345                this=seq_get(args, 0), expression=seq_get(args, 1) or exp.Literal.string("$")
346            ),
347            "MD5": exp.MD5Digest.from_arg_list,
348            "TO_HEX": _build_to_hex,
349            "PARSE_DATE": lambda args: build_formatted_time(exp.StrToDate, "bigquery")(
350                [seq_get(args, 1), seq_get(args, 0)]
351            ),
352            "PARSE_TIMESTAMP": _build_parse_timestamp,
353            "REGEXP_CONTAINS": exp.RegexpLike.from_arg_list,
354            "REGEXP_EXTRACT": lambda args: exp.RegexpExtract(
355                this=seq_get(args, 0),
356                expression=seq_get(args, 1),
357                position=seq_get(args, 2),
358                occurrence=seq_get(args, 3),
359                group=exp.Literal.number(1) if re.compile(args[1].name).groups == 1 else None,
360            ),
361            "SHA256": lambda args: exp.SHA2(this=seq_get(args, 0), length=exp.Literal.number(256)),
362            "SHA512": lambda args: exp.SHA2(this=seq_get(args, 0), length=exp.Literal.number(512)),
363            "SPLIT": lambda args: exp.Split(
364                # https://cloud.google.com/bigquery/docs/reference/standard-sql/string_functions#split
365                this=seq_get(args, 0),
366                expression=seq_get(args, 1) or exp.Literal.string(","),
367            ),
368            "TIME": _build_time,
369            "TIME_ADD": build_date_delta_with_interval(exp.TimeAdd),
370            "TIME_SUB": build_date_delta_with_interval(exp.TimeSub),
371            "TIMESTAMP": _build_timestamp,
372            "TIMESTAMP_ADD": build_date_delta_with_interval(exp.TimestampAdd),
373            "TIMESTAMP_SUB": build_date_delta_with_interval(exp.TimestampSub),
374            "TIMESTAMP_MICROS": lambda args: exp.UnixToTime(
375                this=seq_get(args, 0), scale=exp.UnixToTime.MICROS
376            ),
377            "TIMESTAMP_MILLIS": lambda args: exp.UnixToTime(
378                this=seq_get(args, 0), scale=exp.UnixToTime.MILLIS
379            ),
380            "TIMESTAMP_SECONDS": lambda args: exp.UnixToTime(this=seq_get(args, 0)),
381            "TO_JSON_STRING": exp.JSONFormat.from_arg_list,
382        }
383
384        FUNCTION_PARSERS = {
385            **parser.Parser.FUNCTION_PARSERS,
386            "ARRAY": lambda self: self.expression(exp.Array, expressions=[self._parse_statement()]),
387        }
388        FUNCTION_PARSERS.pop("TRIM")
389
390        NO_PAREN_FUNCTIONS = {
391            **parser.Parser.NO_PAREN_FUNCTIONS,
392            TokenType.CURRENT_DATETIME: exp.CurrentDatetime,
393        }
394
395        NESTED_TYPE_TOKENS = {
396            *parser.Parser.NESTED_TYPE_TOKENS,
397            TokenType.TABLE,
398        }
399
400        PROPERTY_PARSERS = {
401            **parser.Parser.PROPERTY_PARSERS,
402            "NOT DETERMINISTIC": lambda self: self.expression(
403                exp.StabilityProperty, this=exp.Literal.string("VOLATILE")
404            ),
405            "OPTIONS": lambda self: self._parse_with_property(),
406        }
407
408        CONSTRAINT_PARSERS = {
409            **parser.Parser.CONSTRAINT_PARSERS,
410            "OPTIONS": lambda self: exp.Properties(expressions=self._parse_with_property()),
411        }
412
413        RANGE_PARSERS = parser.Parser.RANGE_PARSERS.copy()
414        RANGE_PARSERS.pop(TokenType.OVERLAPS)
415
416        NULL_TOKENS = {TokenType.NULL, TokenType.UNKNOWN}
417
418        STATEMENT_PARSERS = {
419            **parser.Parser.STATEMENT_PARSERS,
420            TokenType.ELSE: lambda self: self._parse_as_command(self._prev),
421            TokenType.END: lambda self: self._parse_as_command(self._prev),
422            TokenType.FOR: lambda self: self._parse_for_in(),
423        }
424
425        BRACKET_OFFSETS = {
426            "OFFSET": (0, False),
427            "ORDINAL": (1, False),
428            "SAFE_OFFSET": (0, True),
429            "SAFE_ORDINAL": (1, True),
430        }
431
432        def _parse_for_in(self) -> exp.ForIn:
433            this = self._parse_range()
434            self._match_text_seq("DO")
435            return self.expression(exp.ForIn, this=this, expression=self._parse_statement())
436
437        def _parse_table_part(self, schema: bool = False) -> t.Optional[exp.Expression]:
438            this = super()._parse_table_part(schema=schema) or self._parse_number()
439
440            # https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#table_names
441            if isinstance(this, exp.Identifier):
442                table_name = this.name
443                while self._match(TokenType.DASH, advance=False) and self._next:
444                    text = ""
445                    while self._curr and self._curr.token_type != TokenType.DOT:
446                        self._advance()
447                        text += self._prev.text
448                    table_name += text
449
450                this = exp.Identifier(this=table_name, quoted=this.args.get("quoted"))
451            elif isinstance(this, exp.Literal):
452                table_name = this.name
453
454                if self._is_connected() and self._parse_var(any_token=True):
455                    table_name += self._prev.text
456
457                this = exp.Identifier(this=table_name, quoted=True)
458
459            return this
460
461        def _parse_table_parts(
462            self, schema: bool = False, is_db_reference: bool = False, wildcard: bool = False
463        ) -> exp.Table:
464            table = super()._parse_table_parts(
465                schema=schema, is_db_reference=is_db_reference, wildcard=True
466            )
467
468            # proj-1.db.tbl -- `1.` is tokenized as a float so we need to unravel it here
469            if not table.catalog:
470                if table.db:
471                    parts = table.db.split(".")
472                    if len(parts) == 2 and not table.args["db"].quoted:
473                        table.set("catalog", exp.Identifier(this=parts[0]))
474                        table.set("db", exp.Identifier(this=parts[1]))
475                else:
476                    parts = table.name.split(".")
477                    if len(parts) == 2 and not table.this.quoted:
478                        table.set("db", exp.Identifier(this=parts[0]))
479                        table.set("this", exp.Identifier(this=parts[1]))
480
481            if any("." in p.name for p in table.parts):
482                catalog, db, this, *rest = (
483                    exp.to_identifier(p, quoted=True)
484                    for p in split_num_words(".".join(p.name for p in table.parts), ".", 3)
485                )
486
487                if rest and this:
488                    this = exp.Dot.build([this, *rest])  # type: ignore
489
490                table = exp.Table(
491                    this=this, db=db, catalog=catalog, pivots=table.args.get("pivots")
492                )
493                table.meta["quoted_table"] = True
494
495            return table
496
497        def _parse_column(self) -> t.Optional[exp.Expression]:
498            column = super()._parse_column()
499            if isinstance(column, exp.Column):
500                parts = column.parts
501                if any("." in p.name for p in parts):
502                    catalog, db, table, this, *rest = (
503                        exp.to_identifier(p, quoted=True)
504                        for p in split_num_words(".".join(p.name for p in parts), ".", 4)
505                    )
506
507                    if rest and this:
508                        this = exp.Dot.build([this, *rest])  # type: ignore
509
510                    column = exp.Column(this=this, table=table, db=db, catalog=catalog)
511                    column.meta["quoted_column"] = True
512
513            return column
514
515        @t.overload
516        def _parse_json_object(self, agg: Lit[False]) -> exp.JSONObject: ...
517
518        @t.overload
519        def _parse_json_object(self, agg: Lit[True]) -> exp.JSONObjectAgg: ...
520
521        def _parse_json_object(self, agg=False):
522            json_object = super()._parse_json_object()
523            array_kv_pair = seq_get(json_object.expressions, 0)
524
525            # Converts BQ's "signature 2" of JSON_OBJECT into SQLGlot's canonical representation
526            # https://cloud.google.com/bigquery/docs/reference/standard-sql/json_functions#json_object_signature2
527            if (
528                array_kv_pair
529                and isinstance(array_kv_pair.this, exp.Array)
530                and isinstance(array_kv_pair.expression, exp.Array)
531            ):
532                keys = array_kv_pair.this.expressions
533                values = array_kv_pair.expression.expressions
534
535                json_object.set(
536                    "expressions",
537                    [exp.JSONKeyValue(this=k, expression=v) for k, v in zip(keys, values)],
538                )
539
540            return json_object
541
542        def _parse_bracket(
543            self, this: t.Optional[exp.Expression] = None
544        ) -> t.Optional[exp.Expression]:
545            bracket = super()._parse_bracket(this)
546
547            if this is bracket:
548                return bracket
549
550            if isinstance(bracket, exp.Bracket):
551                for expression in bracket.expressions:
552                    name = expression.name.upper()
553
554                    if name not in self.BRACKET_OFFSETS:
555                        break
556
557                    offset, safe = self.BRACKET_OFFSETS[name]
558                    bracket.set("offset", offset)
559                    bracket.set("safe", safe)
560                    expression.replace(expression.expressions[0])
561
562            return bracket

Parser consumes a list of tokens produced by the Tokenizer and produces a parsed syntax tree.

Arguments:
  • error_level: The desired error level. Default: ErrorLevel.IMMEDIATE
  • error_message_context: The amount of context to capture from a query string when displaying the error message (in number of characters). Default: 100
  • max_errors: Maximum number of error messages to include in a raised ParseError. This is only relevant if error_level is ErrorLevel.RAISE. Default: 3
PREFIXED_PIVOT_COLUMNS = True
LOG_DEFAULTS_TO_LN = True
SUPPORTS_IMPLICIT_UNNEST = True
FUNCTIONS = {'ABS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Abs'>>, 'ADD_MONTHS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.AddMonths'>>, 'ANONYMOUS_AGG_FUNC': <bound method Func.from_arg_list of <class 'sqlglot.expressions.AnonymousAggFunc'>>, 'ANY_VALUE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.AnyValue'>>, 'APPROX_DISTINCT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ApproxDistinct'>>, 'APPROX_COUNT_DISTINCT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ApproxDistinct'>>, 'APPROX_QUANTILE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ApproxQuantile'>>, 'APPROX_TOP_K': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ApproxTopK'>>, 'ARG_MAX': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArgMax'>>, 'ARGMAX': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArgMax'>>, 'MAX_BY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArgMax'>>, 'ARG_MIN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArgMin'>>, 'ARGMIN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArgMin'>>, 'MIN_BY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArgMin'>>, 'ARRAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Array'>>, 'ARRAY_AGG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayAgg'>>, 'ARRAY_ALL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayAll'>>, 'ARRAY_ANY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayAny'>>, 'ARRAY_CONCAT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayConcat'>>, 'ARRAY_CAT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayConcat'>>, 'ARRAY_CONSTRUCT_COMPACT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayConstructCompact'>>, 'ARRAY_CONTAINS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayContains'>>, 'ARRAY_HAS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayContains'>>, 'ARRAY_CONTAINS_ALL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayContainsAll'>>, 'ARRAY_HAS_ALL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayContainsAll'>>, 'FILTER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayFilter'>>, 'ARRAY_FILTER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayFilter'>>, 'ARRAY_OVERLAPS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayOverlaps'>>, 'ARRAY_SIZE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArraySize'>>, 'ARRAY_LENGTH': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArraySize'>>, 'ARRAY_SORT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArraySort'>>, 'ARRAY_SUM': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArraySum'>>, 'ARRAY_TO_STRING': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayToString'>>, 'ARRAY_JOIN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayToString'>>, 'ARRAY_UNION_AGG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayUnionAgg'>>, 'ARRAY_UNIQUE_AGG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayUniqueAgg'>>, 'AVG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Avg'>>, 'CASE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Case'>>, 'CAST': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Cast'>>, 'CAST_TO_STR_TYPE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CastToStrType'>>, 'CBRT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Cbrt'>>, 'CEIL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Ceil'>>, 'CEILING': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Ceil'>>, 'CHR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Chr'>>, 'CHAR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Chr'>>, 'COALESCE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Coalesce'>>, 'IFNULL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Coalesce'>>, 'NVL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Coalesce'>>, 'COLLATE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Collate'>>, 'COMBINED_AGG_FUNC': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CombinedAggFunc'>>, 'COMBINED_PARAMETERIZED_AGG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CombinedParameterizedAgg'>>, 'CONCAT': <function Parser.<lambda>>, 'CONCAT_WS': <function Parser.<lambda>>, 'CONNECT_BY_ROOT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ConnectByRoot'>>, 'CONVERT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Convert'>>, 'CORR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Corr'>>, 'COUNT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Count'>>, 'COUNT_IF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CountIf'>>, 'COUNTIF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CountIf'>>, 'COVAR_POP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CovarPop'>>, 'COVAR_SAMP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CovarSamp'>>, 'CURRENT_DATE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CurrentDate'>>, 'CURRENT_DATETIME': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CurrentDatetime'>>, 'CURRENT_TIME': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CurrentTime'>>, 'CURRENT_TIMESTAMP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CurrentTimestamp'>>, 'CURRENT_USER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CurrentUser'>>, 'DATE': <function _build_date>, 'DATE_ADD': <function build_date_delta_with_interval.<locals>._builder>, 'DATEDIFF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DateDiff'>>, 'DATE_DIFF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DateDiff'>>, 'DATE_FROM_PARTS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DateFromParts'>>, 'DATEFROMPARTS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DateFromParts'>>, 'DATE_STR_TO_DATE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DateStrToDate'>>, 'DATE_SUB': <function build_date_delta_with_interval.<locals>._builder>, 'DATE_TO_DATE_STR': <function Parser.<lambda>>, 'DATE_TO_DI': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DateToDi'>>, 'DATE_TRUNC': <function BigQuery.Parser.<lambda>>, 'DATETIME': <function _build_datetime>, 'DATETIME_ADD': <function build_date_delta_with_interval.<locals>._builder>, 'DATETIME_DIFF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DatetimeDiff'>>, 'DATETIME_SUB': <function build_date_delta_with_interval.<locals>._builder>, 'DATETIME_TRUNC': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DatetimeTrunc'>>, 'DAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Day'>>, 'DAY_OF_MONTH': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DayOfMonth'>>, 'DAYOFMONTH': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DayOfMonth'>>, 'DAY_OF_WEEK': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DayOfWeek'>>, 'DAYOFWEEK': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DayOfWeek'>>, 'DAY_OF_YEAR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DayOfYear'>>, 'DAYOFYEAR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DayOfYear'>>, 'DECODE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Decode'>>, 'DI_TO_DATE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DiToDate'>>, 'ENCODE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Encode'>>, 'EXP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Exp'>>, 'EXPLODE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Explode'>>, 'EXPLODE_OUTER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ExplodeOuter'>>, 'EXTRACT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Extract'>>, 'FIRST': <bound method Func.from_arg_list of <class 'sqlglot.expressions.First'>>, 'FIRST_VALUE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.FirstValue'>>, 'FLATTEN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Flatten'>>, 'FLOOR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Floor'>>, 'FROM_BASE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.FromBase'>>, 'FROM_BASE64': <bound method Func.from_arg_list of <class 'sqlglot.expressions.FromBase64'>>, 'GAP_FILL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.GapFill'>>, 'GENERATE_DATE_ARRAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.GenerateDateArray'>>, 'GENERATE_SERIES': <bound method Func.from_arg_list of <class 'sqlglot.expressions.GenerateSeries'>>, 'GREATEST': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Greatest'>>, 'GROUP_CONCAT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.GroupConcat'>>, 'HEX': <function build_hex>, 'HLL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Hll'>>, 'IF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.If'>>, 'IIF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.If'>>, 'INITCAP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Initcap'>>, 'IS_INF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.IsInf'>>, 'ISINF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.IsInf'>>, 'IS_NAN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.IsNan'>>, 'ISNAN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.IsNan'>>, 'J_S_O_N_ARRAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONArray'>>, 'J_S_O_N_ARRAY_AGG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONArrayAgg'>>, 'JSON_ARRAY_CONTAINS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONArrayContains'>>, 'JSONB_CONTAINS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONBContains'>>, 'JSONB_EXTRACT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONBExtract'>>, 'JSONB_EXTRACT_SCALAR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONBExtractScalar'>>, 'JSON_EXTRACT': <function build_extract_json_with_path.<locals>._builder>, 'JSON_EXTRACT_SCALAR': <function BigQuery.Parser.<lambda>>, 'JSON_FORMAT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONFormat'>>, 'J_S_O_N_OBJECT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONObject'>>, 'J_S_O_N_OBJECT_AGG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONObjectAgg'>>, 'J_S_O_N_TABLE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONTable'>>, 'LAG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Lag'>>, 'LAST': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Last'>>, 'LAST_DAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LastDay'>>, 'LAST_DAY_OF_MONTH': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LastDay'>>, 'LAST_VALUE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LastValue'>>, 'LEAD': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Lead'>>, 'LEAST': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Least'>>, 'LEFT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Left'>>, 'LENGTH': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Length'>>, 'LEN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Length'>>, 'LEVENSHTEIN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Levenshtein'>>, 'LIST': <bound method Func.from_arg_list of <class 'sqlglot.expressions.List'>>, 'LN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Ln'>>, 'LOG': <function build_logarithm>, 'LOGICAL_AND': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LogicalAnd'>>, 'BOOL_AND': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LogicalAnd'>>, 'BOOLAND_AGG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LogicalAnd'>>, 'LOGICAL_OR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LogicalOr'>>, 'BOOL_OR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LogicalOr'>>, 'BOOLOR_AGG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LogicalOr'>>, 'LOWER': <function build_lower>, 'LCASE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Lower'>>, 'LOWER_HEX': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LowerHex'>>, 'MD5': <bound method Func.from_arg_list of <class 'sqlglot.expressions.MD5Digest'>>, 'MD5_DIGEST': <bound method Func.from_arg_list of <class 'sqlglot.expressions.MD5Digest'>>, 'MAP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Map'>>, 'MAP_FROM_ENTRIES': <bound method Func.from_arg_list of <class 'sqlglot.expressions.MapFromEntries'>>, 'MATCH_AGAINST': <bound method Func.from_arg_list of <class 'sqlglot.expressions.MatchAgainst'>>, 'MAX': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Max'>>, 'MIN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Min'>>, 'MONTH': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Month'>>, 'MONTHS_BETWEEN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.MonthsBetween'>>, 'NEXT_VALUE_FOR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.NextValueFor'>>, 'NTH_VALUE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.NthValue'>>, 'NULLIF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Nullif'>>, 'NUMBER_TO_STR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.NumberToStr'>>, 'NVL2': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Nvl2'>>, 'OPEN_J_S_O_N': <bound method Func.from_arg_list of <class 'sqlglot.expressions.OpenJSON'>>, 'PARAMETERIZED_AGG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ParameterizedAgg'>>, 'PARSE_JSON': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ParseJSON'>>, 'JSON_PARSE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ParseJSON'>>, 'PERCENTILE_CONT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.PercentileCont'>>, 'PERCENTILE_DISC': <bound method Func.from_arg_list of <class 'sqlglot.expressions.PercentileDisc'>>, 'POSEXPLODE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Posexplode'>>, 'POSEXPLODE_OUTER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.PosexplodeOuter'>>, 'POWER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Pow'>>, 'POW': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Pow'>>, 'PREDICT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Predict'>>, 'QUANTILE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Quantile'>>, 'QUARTER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Quarter'>>, 'RAND': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Rand'>>, 'RANDOM': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Rand'>>, 'RANDN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Randn'>>, 'RANGE_N': <bound method Func.from_arg_list of <class 'sqlglot.expressions.RangeN'>>, 'READ_CSV': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ReadCSV'>>, 'REDUCE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Reduce'>>, 'REGEXP_EXTRACT': <function BigQuery.Parser.<lambda>>, 'REGEXP_I_LIKE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.RegexpILike'>>, 'REGEXP_LIKE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.RegexpLike'>>, 'REGEXP_REPLACE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.RegexpReplace'>>, 'REGEXP_SPLIT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.RegexpSplit'>>, 'REPEAT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Repeat'>>, 'RIGHT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Right'>>, 'ROUND': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Round'>>, 'ROW_NUMBER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.RowNumber'>>, 'SHA': <bound method Func.from_arg_list of <class 'sqlglot.expressions.SHA'>>, 'SHA1': <bound method Func.from_arg_list of <class 'sqlglot.expressions.SHA'>>, 'SHA2': <bound method Func.from_arg_list of <class 'sqlglot.expressions.SHA2'>>, 'SAFE_DIVIDE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.SafeDivide'>>, 'SIGN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Sign'>>, 'SIGNUM': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Sign'>>, 'SORT_ARRAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.SortArray'>>, 'SPLIT': <function BigQuery.Parser.<lambda>>, 'SQRT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Sqrt'>>, 'STANDARD_HASH': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StandardHash'>>, 'STAR_MAP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StarMap'>>, 'STARTS_WITH': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StartsWith'>>, 'STARTSWITH': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StartsWith'>>, 'STDDEV': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Stddev'>>, 'STDDEV_POP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StddevPop'>>, 'STDDEV_SAMP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StddevSamp'>>, 'STR_POSITION': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StrPosition'>>, 'STR_TO_DATE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StrToDate'>>, 'STR_TO_MAP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StrToMap'>>, 'STR_TO_TIME': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StrToTime'>>, 'STR_TO_UNIX': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StrToUnix'>>, 'STRING_TO_ARRAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StringToArray'>>, 'SPLIT_BY_STRING': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StringToArray'>>, 'STRUCT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Struct'>>, 'STRUCT_EXTRACT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StructExtract'>>, 'STUFF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Stuff'>>, 'INSERT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Stuff'>>, 'SUBSTRING': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Substring'>>, 'SUM': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Sum'>>, 'TIME': <function _build_time>, 'TIME_ADD': <function build_date_delta_with_interval.<locals>._builder>, 'TIME_DIFF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TimeDiff'>>, 'TIME_FROM_PARTS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TimeFromParts'>>, 'TIMEFROMPARTS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TimeFromParts'>>, 'TIME_STR_TO_DATE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TimeStrToDate'>>, 'TIME_STR_TO_TIME': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TimeStrToTime'>>, 'TIME_STR_TO_UNIX': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TimeStrToUnix'>>, 'TIME_SUB': <function build_date_delta_with_interval.<locals>._builder>, 'TIME_TO_STR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TimeToStr'>>, 'TIME_TO_TIME_STR': <function Parser.<lambda>>, 'TIME_TO_UNIX': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TimeToUnix'>>, 'TIME_TRUNC': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TimeTrunc'>>, 'TIMESTAMP': <function _build_timestamp>, 'TIMESTAMP_ADD': <function build_date_delta_with_interval.<locals>._builder>, 'TIMESTAMPDIFF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TimestampDiff'>>, 'TIMESTAMP_DIFF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TimestampDiff'>>, 'TIMESTAMP_FROM_PARTS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TimestampFromParts'>>, 'TIMESTAMPFROMPARTS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TimestampFromParts'>>, 'TIMESTAMP_SUB': <function build_date_delta_with_interval.<locals>._builder>, 'TIMESTAMP_TRUNC': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TimestampTrunc'>>, 'TO_ARRAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ToArray'>>, 'TO_BASE64': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ToBase64'>>, 'TO_CHAR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ToChar'>>, 'TO_DAYS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ToDays'>>, 'TO_MAP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ToMap'>>, 'TO_NUMBER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ToNumber'>>, 'TRANSFORM': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Transform'>>, 'TRIM': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Trim'>>, 'TRY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Try'>>, 'TRY_CAST': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TryCast'>>, 'TS_OR_DI_TO_DI': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDiToDi'>>, 'TS_OR_DS_ADD': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDsAdd'>>, 'TS_OR_DS_DIFF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDsDiff'>>, 'TS_OR_DS_TO_DATE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDsToDate'>>, 'TS_OR_DS_TO_DATE_STR': <function Parser.<lambda>>, 'TS_OR_DS_TO_TIME': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDsToTime'>>, 'TS_OR_DS_TO_TIMESTAMP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDsToTimestamp'>>, 'UNHEX': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Unhex'>>, 'UNIX_DATE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.UnixDate'>>, 'UNIX_TO_STR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.UnixToStr'>>, 'UNIX_TO_TIME': <bound method Func.from_arg_list of <class 'sqlglot.expressions.UnixToTime'>>, 'UNIX_TO_TIME_STR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.UnixToTimeStr'>>, 'UNNEST': <function Parser.<lambda>>, 'UPPER': <function build_upper>, 'UCASE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Upper'>>, 'VAR_MAP': <function build_var_map>, 'VARIANCE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Variance'>>, 'VARIANCE_SAMP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Variance'>>, 'VAR_SAMP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Variance'>>, 'VARIANCE_POP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.VariancePop'>>, 'VAR_POP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.VariancePop'>>, 'WEEK': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Week'>>, 'WEEK_OF_YEAR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.WeekOfYear'>>, 'WEEKOFYEAR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.WeekOfYear'>>, 'WHEN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.When'>>, 'X_M_L_TABLE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.XMLTable'>>, 'XOR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Xor'>>, 'YEAR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Year'>>, 'GLOB': <function Parser.<lambda>>, 'JSON_EXTRACT_PATH_TEXT': <function build_extract_json_with_path.<locals>._builder>, 'LIKE': <function build_like>, 'LOG2': <function Parser.<lambda>>, 'LOG10': <function Parser.<lambda>>, 'MOD': <function build_mod>, 'TO_HEX': <function _build_to_hex>, 'DIV': <function binary_from_function.<locals>.<lambda>>, 'FORMAT_DATE': <function BigQuery.Parser.<lambda>>, 'GENERATE_ARRAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.GenerateSeries'>>, 'PARSE_DATE': <function BigQuery.Parser.<lambda>>, 'PARSE_TIMESTAMP': <function _build_parse_timestamp>, 'REGEXP_CONTAINS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.RegexpLike'>>, 'SHA256': <function BigQuery.Parser.<lambda>>, 'SHA512': <function BigQuery.Parser.<lambda>>, 'TIMESTAMP_MICROS': <function BigQuery.Parser.<lambda>>, 'TIMESTAMP_MILLIS': <function BigQuery.Parser.<lambda>>, 'TIMESTAMP_SECONDS': <function BigQuery.Parser.<lambda>>, 'TO_JSON_STRING': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONFormat'>>}
FUNCTION_PARSERS = {'CAST': <function Parser.<lambda>>, 'CONVERT': <function Parser.<lambda>>, 'DECODE': <function Parser.<lambda>>, 'EXTRACT': <function Parser.<lambda>>, 'GAP_FILL': <function Parser.<lambda>>, 'JSON_OBJECT': <function Parser.<lambda>>, 'JSON_OBJECTAGG': <function Parser.<lambda>>, 'JSON_TABLE': <function Parser.<lambda>>, 'MATCH': <function Parser.<lambda>>, 'OPENJSON': <function Parser.<lambda>>, 'POSITION': <function Parser.<lambda>>, 'PREDICT': <function Parser.<lambda>>, 'SAFE_CAST': <function Parser.<lambda>>, 'STRING_AGG': <function Parser.<lambda>>, 'SUBSTRING': <function Parser.<lambda>>, 'TRY_CAST': <function Parser.<lambda>>, 'TRY_CONVERT': <function Parser.<lambda>>, 'ARRAY': <function BigQuery.Parser.<lambda>>}
NO_PAREN_FUNCTIONS = {<TokenType.CURRENT_DATE: 'CURRENT_DATE'>: <class 'sqlglot.expressions.CurrentDate'>, <TokenType.CURRENT_DATETIME: 'CURRENT_DATETIME'>: <class 'sqlglot.expressions.CurrentDatetime'>, <TokenType.CURRENT_TIME: 'CURRENT_TIME'>: <class 'sqlglot.expressions.CurrentTime'>, <TokenType.CURRENT_TIMESTAMP: 'CURRENT_TIMESTAMP'>: <class 'sqlglot.expressions.CurrentTimestamp'>, <TokenType.CURRENT_USER: 'CURRENT_USER'>: <class 'sqlglot.expressions.CurrentUser'>}
NESTED_TYPE_TOKENS = {<TokenType.MAP: 'MAP'>, <TokenType.ARRAY: 'ARRAY'>, <TokenType.OBJECT: 'OBJECT'>, <TokenType.NESTED: 'NESTED'>, <TokenType.NULLABLE: 'NULLABLE'>, <TokenType.LOWCARDINALITY: 'LOWCARDINALITY'>, <TokenType.TABLE: 'TABLE'>, <TokenType.LIST: 'LIST'>, <TokenType.STRUCT: 'STRUCT'>}
PROPERTY_PARSERS = {'ALLOWED_VALUES': <function Parser.<lambda>>, 'ALGORITHM': <function Parser.<lambda>>, 'AUTO': <function Parser.<lambda>>, 'AUTO_INCREMENT': <function Parser.<lambda>>, 'BACKUP': <function Parser.<lambda>>, 'BLOCKCOMPRESSION': <function Parser.<lambda>>, 'CHARSET': <function Parser.<lambda>>, 'CHARACTER SET': <function Parser.<lambda>>, 'CHECKSUM': <function Parser.<lambda>>, 'CLUSTER BY': <function Parser.<lambda>>, 'CLUSTERED': <function Parser.<lambda>>, 'COLLATE': <function Parser.<lambda>>, 'COMMENT': <function Parser.<lambda>>, 'CONTAINS': <function Parser.<lambda>>, 'COPY': <function Parser.<lambda>>, 'DATABLOCKSIZE': <function Parser.<lambda>>, 'DATA_DELETION': <function Parser.<lambda>>, 'DEFINER': <function Parser.<lambda>>, 'DETERMINISTIC': <function Parser.<lambda>>, 'DISTKEY': <function Parser.<lambda>>, 'DISTSTYLE': <function Parser.<lambda>>, 'ENGINE': <function Parser.<lambda>>, 'EXECUTE': <function Parser.<lambda>>, 'EXTERNAL': <function Parser.<lambda>>, 'FALLBACK': <function Parser.<lambda>>, 'FORMAT': <function Parser.<lambda>>, 'FREESPACE': <function Parser.<lambda>>, 'GLOBAL': <function Parser.<lambda>>, 'HEAP': <function Parser.<lambda>>, 'ICEBERG': <function Parser.<lambda>>, 'IMMUTABLE': <function Parser.<lambda>>, 'INHERITS': <function Parser.<lambda>>, 'INPUT': <function Parser.<lambda>>, 'JOURNAL': <function Parser.<lambda>>, 'LANGUAGE': <function Parser.<lambda>>, 'LAYOUT': <function Parser.<lambda>>, 'LIFETIME': <function Parser.<lambda>>, 'LIKE': <function Parser.<lambda>>, 'LOCATION': <function Parser.<lambda>>, 'LOCK': <function Parser.<lambda>>, 'LOCKING': <function Parser.<lambda>>, 'LOG': <function Parser.<lambda>>, 'MATERIALIZED': <function Parser.<lambda>>, 'MERGEBLOCKRATIO': <function Parser.<lambda>>, 'MODIFIES': <function Parser.<lambda>>, 'MULTISET': <function Parser.<lambda>>, 'NO': <function Parser.<lambda>>, 'ON': <function Parser.<lambda>>, 'ORDER BY': <function Parser.<lambda>>, 'OUTPUT': <function Parser.<lambda>>, 'PARTITION': <function Parser.<lambda>>, 'PARTITION BY': <function Parser.<lambda>>, 'PARTITIONED BY': <function Parser.<lambda>>, 'PARTITIONED_BY': <function Parser.<lambda>>, 'PRIMARY KEY': <function Parser.<lambda>>, 'RANGE': <function Parser.<lambda>>, 'READS': <function Parser.<lambda>>, 'REMOTE': <function Parser.<lambda>>, 'RETURNS': <function Parser.<lambda>>, 'STRICT': <function Parser.<lambda>>, 'ROW': <function Parser.<lambda>>, 'ROW_FORMAT': <function Parser.<lambda>>, 'SAMPLE': <function Parser.<lambda>>, 'SECURE': <function Parser.<lambda>>, 'SET': <function Parser.<lambda>>, 'SETTINGS': <function Parser.<lambda>>, 'SHARING': <function Parser.<lambda>>, 'SORTKEY': <function Parser.<lambda>>, 'SOURCE': <function Parser.<lambda>>, 'STABLE': <function Parser.<lambda>>, 'STORED': <function Parser.<lambda>>, 'SYSTEM_VERSIONING': <function Parser.<lambda>>, 'TBLPROPERTIES': <function Parser.<lambda>>, 'TEMP': <function Parser.<lambda>>, 'TEMPORARY': <function Parser.<lambda>>, 'TO': <function Parser.<lambda>>, 'TRANSIENT': <function Parser.<lambda>>, 'TRANSFORM': <function Parser.<lambda>>, 'TTL': <function Parser.<lambda>>, 'USING': <function Parser.<lambda>>, 'UNLOGGED': <function Parser.<lambda>>, 'VOLATILE': <function Parser.<lambda>>, 'WITH': <function Parser.<lambda>>, 'NOT DETERMINISTIC': <function BigQuery.Parser.<lambda>>, 'OPTIONS': <function BigQuery.Parser.<lambda>>}
CONSTRAINT_PARSERS = {'AUTOINCREMENT': <function Parser.<lambda>>, 'AUTO_INCREMENT': <function Parser.<lambda>>, 'CASESPECIFIC': <function Parser.<lambda>>, 'CHARACTER SET': <function Parser.<lambda>>, 'CHECK': <function Parser.<lambda>>, 'COLLATE': <function Parser.<lambda>>, 'COMMENT': <function Parser.<lambda>>, 'COMPRESS': <function Parser.<lambda>>, 'CLUSTERED': <function Parser.<lambda>>, 'NONCLUSTERED': <function Parser.<lambda>>, 'DEFAULT': <function Parser.<lambda>>, 'ENCODE': <function Parser.<lambda>>, 'EPHEMERAL': <function Parser.<lambda>>, 'EXCLUDE': <function Parser.<lambda>>, 'FOREIGN KEY': <function Parser.<lambda>>, 'FORMAT': <function Parser.<lambda>>, 'GENERATED': <function Parser.<lambda>>, 'IDENTITY': <function Parser.<lambda>>, 'INLINE': <function Parser.<lambda>>, 'LIKE': <function Parser.<lambda>>, 'NOT': <function Parser.<lambda>>, 'NULL': <function Parser.<lambda>>, 'ON': <function Parser.<lambda>>, 'PATH': <function Parser.<lambda>>, 'PERIOD': <function Parser.<lambda>>, 'PRIMARY KEY': <function Parser.<lambda>>, 'REFERENCES': <function Parser.<lambda>>, 'TITLE': <function Parser.<lambda>>, 'TTL': <function Parser.<lambda>>, 'UNIQUE': <function Parser.<lambda>>, 'UPPERCASE': <function Parser.<lambda>>, 'WITH': <function Parser.<lambda>>, 'OPTIONS': <function BigQuery.Parser.<lambda>>}
RANGE_PARSERS = {<TokenType.BETWEEN: 'BETWEEN'>: <function Parser.<lambda>>, <TokenType.GLOB: 'GLOB'>: <function binary_range_parser.<locals>._parse_binary_range>, <TokenType.ILIKE: 'ILIKE'>: <function binary_range_parser.<locals>._parse_binary_range>, <TokenType.IN: 'IN'>: <function Parser.<lambda>>, <TokenType.IRLIKE: 'IRLIKE'>: <function binary_range_parser.<locals>._parse_binary_range>, <TokenType.IS: 'IS'>: <function Parser.<lambda>>, <TokenType.LIKE: 'LIKE'>: <function binary_range_parser.<locals>._parse_binary_range>, <TokenType.RLIKE: 'RLIKE'>: <function binary_range_parser.<locals>._parse_binary_range>, <TokenType.SIMILAR_TO: 'SIMILAR_TO'>: <function binary_range_parser.<locals>._parse_binary_range>, <TokenType.FOR: 'FOR'>: <function Parser.<lambda>>}
NULL_TOKENS = {<TokenType.UNKNOWN: 'UNKNOWN'>, <TokenType.NULL: 'NULL'>}
STATEMENT_PARSERS = {<TokenType.ALTER: 'ALTER'>: <function Parser.<lambda>>, <TokenType.BEGIN: 'BEGIN'>: <function Parser.<lambda>>, <TokenType.CACHE: 'CACHE'>: <function Parser.<lambda>>, <TokenType.COMMENT: 'COMMENT'>: <function Parser.<lambda>>, <TokenType.COMMIT: 'COMMIT'>: <function Parser.<lambda>>, <TokenType.COPY: 'COPY'>: <function Parser.<lambda>>, <TokenType.CREATE: 'CREATE'>: <function Parser.<lambda>>, <TokenType.DELETE: 'DELETE'>: <function Parser.<lambda>>, <TokenType.DESC: 'DESC'>: <function Parser.<lambda>>, <TokenType.DESCRIBE: 'DESCRIBE'>: <function Parser.<lambda>>, <TokenType.DROP: 'DROP'>: <function Parser.<lambda>>, <TokenType.INSERT: 'INSERT'>: <function Parser.<lambda>>, <TokenType.KILL: 'KILL'>: <function Parser.<lambda>>, <TokenType.LOAD: 'LOAD'>: <function Parser.<lambda>>, <TokenType.MERGE: 'MERGE'>: <function Parser.<lambda>>, <TokenType.PIVOT: 'PIVOT'>: <function Parser.<lambda>>, <TokenType.PRAGMA: 'PRAGMA'>: <function Parser.<lambda>>, <TokenType.REFRESH: 'REFRESH'>: <function Parser.<lambda>>, <TokenType.ROLLBACK: 'ROLLBACK'>: <function Parser.<lambda>>, <TokenType.SET: 'SET'>: <function Parser.<lambda>>, <TokenType.TRUNCATE: 'TRUNCATE'>: <function Parser.<lambda>>, <TokenType.UNCACHE: 'UNCACHE'>: <function Parser.<lambda>>, <TokenType.UPDATE: 'UPDATE'>: <function Parser.<lambda>>, <TokenType.USE: 'USE'>: <function Parser.<lambda>>, <TokenType.SEMICOLON: 'SEMICOLON'>: <function Parser.<lambda>>, <TokenType.ELSE: 'ELSE'>: <function BigQuery.Parser.<lambda>>, <TokenType.END: 'END'>: <function BigQuery.Parser.<lambda>>, <TokenType.FOR: 'FOR'>: <function BigQuery.Parser.<lambda>>}
BRACKET_OFFSETS = {'OFFSET': (0, False), 'ORDINAL': (1, False), 'SAFE_OFFSET': (0, True), 'SAFE_ORDINAL': (1, True)}
ID_VAR_TOKENS = {<TokenType.SERIAL: 'SERIAL'>, <TokenType.TSMULTIRANGE: 'TSMULTIRANGE'>, <TokenType.MAP: 'MAP'>, <TokenType.GEOGRAPHY: 'GEOGRAPHY'>, <TokenType.COMMIT: 'COMMIT'>, <TokenType.NESTED: 'NESTED'>, <TokenType.CURRENT_USER: 'CURRENT_USER'>, <TokenType.DICTIONARY: 'DICTIONARY'>, <TokenType.SMALLINT: 'SMALLINT'>, <TokenType.TABLE: 'TABLE'>, <TokenType.LOAD: 'LOAD'>, <TokenType.EXISTS: 'EXISTS'>, <TokenType.ESCAPE: 'ESCAPE'>, <TokenType.TDIGEST: 'TDIGEST'>, <TokenType.SIMPLEAGGREGATEFUNCTION: 'SIMPLEAGGREGATEFUNCTION'>, <TokenType.DECIMAL: 'DECIMAL'>, <TokenType.INT4MULTIRANGE: 'INT4MULTIRANGE'>, <TokenType.BIT: 'BIT'>, <TokenType.OBJECT_IDENTIFIER: 'OBJECT_IDENTIFIER'>, <TokenType.DATE32: 'DATE32'>, <TokenType.CURRENT_TIME: 'CURRENT_TIME'>, <TokenType.TIME: 'TIME'>, <TokenType.CURRENT_TIMESTAMP: 'CURRENT_TIMESTAMP'>, <TokenType.SMALLSERIAL: 'SMALLSERIAL'>, <TokenType.COMMAND: 'COMMAND'>, <TokenType.TINYBLOB: 'TINYBLOB'>, <TokenType.SUPER: 'SUPER'>, <TokenType.PERCENT: 'PERCENT'>, <TokenType.USMALLINT: 'USMALLINT'>, <TokenType.TRUNCATE: 'TRUNCATE'>, <TokenType.NCHAR: 'NCHAR'>, <TokenType.HLLSKETCH: 'HLLSKETCH'>, <TokenType.CONSTRAINT: 'CONSTRAINT'>, <TokenType.CACHE: 'CACHE'>, <TokenType.BIGSERIAL: 'BIGSERIAL'>, <TokenType.OFFSET: 'OFFSET'>, <TokenType.VIEW: 'VIEW'>, <TokenType.FALSE: 'FALSE'>, <TokenType.WINDOW: 'WINDOW'>, <TokenType.FIRST: 'FIRST'>, <TokenType.JSONB: 'JSONB'>, <TokenType.NULL: 'NULL'>, <TokenType.UINT: 'UINT'>, <TokenType.INET: 'INET'>, <TokenType.UNIQUE: 'UNIQUE'>, <TokenType.SCHEMA: 'SCHEMA'>, <TokenType.DESCRIBE: 'DESCRIBE'>, <TokenType.FILTER: 'FILTER'>, <TokenType.NULLABLE: 'NULLABLE'>, <TokenType.DATETIME: 'DATETIME'>, <TokenType.VARIANT: 'VARIANT'>, <TokenType.USE: 'USE'>, <TokenType.FIXEDSTRING: 'FIXEDSTRING'>, <TokenType.INT: 'INT'>, <TokenType.IPV6: 'IPV6'>, <TokenType.FLOAT: 'FLOAT'>, <TokenType.STORAGE_INTEGRATION: 'STORAGE_INTEGRATION'>, <TokenType.CASE: 'CASE'>, <TokenType.INT8RANGE: 'INT8RANGE'>, <TokenType.DESC: 'DESC'>, <TokenType.CURRENT_DATETIME: 'CURRENT_DATETIME'>, <TokenType.BEGIN: 'BEGIN'>, <TokenType.INT8MULTIRANGE: 'INT8MULTIRANGE'>, <TokenType.KILL: 'KILL'>, <TokenType.TSRANGE: 'TSRANGE'>, <TokenType.IPPREFIX: 'IPPREFIX'>, <TokenType.TINYINT: 'TINYINT'>, <TokenType.STRUCT: 'STRUCT'>, <TokenType.OPERATOR: 'OPERATOR'>, <TokenType.FINAL: 'FINAL'>, <TokenType.REPLACE: 'REPLACE'>, <TokenType.ENUM8: 'ENUM8'>, <TokenType.INT4RANGE: 'INT4RANGE'>, <TokenType.BOOLEAN: 'BOOLEAN'>, <TokenType.DELETE: 'DELETE'>, <TokenType.VAR: 'VAR'>, <TokenType.NVARCHAR: 'NVARCHAR'>, <TokenType.TEXT: 'TEXT'>, <TokenType.TIMESTAMP_NS: 'TIMESTAMP_NS'>, <TokenType.UPDATE: 'UPDATE'>, <TokenType.BIGDECIMAL: 'BIGDECIMAL'>, <TokenType.ROW: 'ROW'>, <TokenType.NAME: 'NAME'>, <TokenType.UINT128: 'UINT128'>, <TokenType.UNIQUEIDENTIFIER: 'UNIQUEIDENTIFIER'>, <TokenType.TRUE: 'TRUE'>, <TokenType.PRAGMA: 'PRAGMA'>, <TokenType.WAREHOUSE: 'WAREHOUSE'>, <TokenType.NUMMULTIRANGE: 'NUMMULTIRANGE'>, <TokenType.JSON: 'JSON'>, <TokenType.OVERLAPS: 'OVERLAPS'>, <TokenType.ALL: 'ALL'>, <TokenType.TIMETZ: 'TIMETZ'>, <TokenType.MEDIUMTEXT: 'MEDIUMTEXT'>, <TokenType.UDECIMAL: 'UDECIMAL'>, <TokenType.ROWS: 'ROWS'>, <TokenType.XML: 'XML'>, <TokenType.UBIGINT: 'UBIGINT'>, <TokenType.PSEUDO_TYPE: 'PSEUDO_TYPE'>, <TokenType.NEXT: 'NEXT'>, <TokenType.UNKNOWN: 'UNKNOWN'>, <TokenType.ROWVERSION: 'ROWVERSION'>, <TokenType.ANY: 'ANY'>, <TokenType.MONEY: 'MONEY'>, <TokenType.TIMESTAMP_MS: 'TIMESTAMP_MS'>, <TokenType.OVERWRITE: 'OVERWRITE'>, <TokenType.OBJECT: 'OBJECT'>, <TokenType.LONGTEXT: 'LONGTEXT'>, <TokenType.UINT256: 'UINT256'>, <TokenType.REFRESH: 'REFRESH'>, <TokenType.LIST: 'LIST'>, <TokenType.STREAMLIT: 'STREAMLIT'>, <TokenType.VARCHAR: 'VARCHAR'>, <TokenType.DATE: 'DATE'>, <TokenType.INT256: 'INT256'>, <TokenType.PIVOT: 'PIVOT'>, <TokenType.TIMESTAMPTZ: 'TIMESTAMPTZ'>, <TokenType.TOP: 'TOP'>, <TokenType.LOWCARDINALITY: 'LOWCARDINALITY'>, <TokenType.CURRENT_DATE: 'CURRENT_DATE'>, <TokenType.SEQUENCE: 'SEQUENCE'>, <TokenType.PROCEDURE: 'PROCEDURE'>, <TokenType.FOREIGN_KEY: 'FOREIGN_KEY'>, <TokenType.ENUM: 'ENUM'>, <TokenType.IDENTIFIER: 'IDENTIFIER'>, <TokenType.FULL: 'FULL'>, <TokenType.MEDIUMINT: 'MEDIUMINT'>, <TokenType.IPADDRESS: 'IPADDRESS'>, <TokenType.ANTI: 'ANTI'>, <TokenType.COPY: 'COPY'>, <TokenType.DATEMULTIRANGE: 'DATEMULTIRANGE'>, <TokenType.VARBINARY: 'VARBINARY'>, <TokenType.YEAR: 'YEAR'>, <TokenType.TSTZRANGE: 'TSTZRANGE'>, <TokenType.DIV: 'DIV'>, <TokenType.FUNCTION: 'FUNCTION'>, <TokenType.COLLATE: 'COLLATE'>, <TokenType.ENUM16: 'ENUM16'>, <TokenType.RANGE: 'RANGE'>, <TokenType.INT128: 'INT128'>, <TokenType.UTINYINT: 'UTINYINT'>, <TokenType.AUTO_INCREMENT: 'AUTO_INCREMENT'>, <TokenType.ASC: 'ASC'>, <TokenType.ROLLUP: 'ROLLUP'>, <TokenType.DATERANGE: 'DATERANGE'>, <TokenType.EXECUTE: 'EXECUTE'>, <TokenType.ASOF: 'ASOF'>, <TokenType.DATABASE: 'DATABASE'>, <TokenType.REFERENCES: 'REFERENCES'>, <TokenType.TSTZMULTIRANGE: 'TSTZMULTIRANGE'>, <TokenType.TIMESTAMPNTZ: 'TIMESTAMPNTZ'>, <TokenType.UNNEST: 'UNNEST'>, <TokenType.DOUBLE: 'DOUBLE'>, <TokenType.TINYTEXT: 'TINYTEXT'>, <TokenType.MERGE: 'MERGE'>, <TokenType.TIMESTAMP_S: 'TIMESTAMP_S'>, <TokenType.USERDEFINED: 'USERDEFINED'>, <TokenType.SEMI: 'SEMI'>, <TokenType.IMAGE: 'IMAGE'>, <TokenType.RIGHT: 'RIGHT'>, <TokenType.COMMENT: 'COMMENT'>, <TokenType.KEEP: 'KEEP'>, <TokenType.HSTORE: 'HSTORE'>, <TokenType.AGGREGATEFUNCTION: 'AGGREGATEFUNCTION'>, <TokenType.NATURAL: 'NATURAL'>, <TokenType.TAG: 'TAG'>, <TokenType.SOME: 'SOME'>, <TokenType.LONGBLOB: 'LONGBLOB'>, <TokenType.MODEL: 'MODEL'>, <TokenType.GEOMETRY: 'GEOMETRY'>, <TokenType.BIGINT: 'BIGINT'>, <TokenType.TIMESTAMP: 'TIMESTAMP'>, <TokenType.SMALLMONEY: 'SMALLMONEY'>, <TokenType.IPV4: 'IPV4'>, <TokenType.RECURSIVE: 'RECURSIVE'>, <TokenType.PARTITION: 'PARTITION'>, <TokenType.APPLY: 'APPLY'>, <TokenType.BINARY: 'BINARY'>, <TokenType.ISNULL: 'ISNULL'>, <TokenType.SHOW: 'SHOW'>, <TokenType.UNPIVOT: 'UNPIVOT'>, <TokenType.SET: 'SET'>, <TokenType.STRAIGHT_JOIN: 'STRAIGHT_JOIN'>, <TokenType.BPCHAR: 'BPCHAR'>, <TokenType.NUMRANGE: 'NUMRANGE'>, <TokenType.MEDIUMBLOB: 'MEDIUMBLOB'>, <TokenType.COLUMN: 'COLUMN'>, <TokenType.END: 'END'>, <TokenType.CHAR: 'CHAR'>, <TokenType.FORMAT: 'FORMAT'>, <TokenType.LEFT: 'LEFT'>, <TokenType.ORDINALITY: 'ORDINALITY'>, <TokenType.INDEX: 'INDEX'>, <TokenType.INTERVAL: 'INTERVAL'>, <TokenType.ARRAY: 'ARRAY'>, <TokenType.UUID: 'UUID'>, <TokenType.DEFAULT: 'DEFAULT'>, <TokenType.TIMESTAMPLTZ: 'TIMESTAMPLTZ'>, <TokenType.DATETIME64: 'DATETIME64'>, <TokenType.IS: 'IS'>, <TokenType.TEMPORARY: 'TEMPORARY'>, <TokenType.VOLATILE: 'VOLATILE'>, <TokenType.SETTINGS: 'SETTINGS'>, <TokenType.UMEDIUMINT: 'UMEDIUMINT'>}
TABLE_ALIAS_TOKENS = {<TokenType.SERIAL: 'SERIAL'>, <TokenType.TSMULTIRANGE: 'TSMULTIRANGE'>, <TokenType.MAP: 'MAP'>, <TokenType.GEOGRAPHY: 'GEOGRAPHY'>, <TokenType.COMMIT: 'COMMIT'>, <TokenType.NESTED: 'NESTED'>, <TokenType.CURRENT_USER: 'CURRENT_USER'>, <TokenType.DICTIONARY: 'DICTIONARY'>, <TokenType.SMALLINT: 'SMALLINT'>, <TokenType.TABLE: 'TABLE'>, <TokenType.LOAD: 'LOAD'>, <TokenType.EXISTS: 'EXISTS'>, <TokenType.ESCAPE: 'ESCAPE'>, <TokenType.TDIGEST: 'TDIGEST'>, <TokenType.SIMPLEAGGREGATEFUNCTION: 'SIMPLEAGGREGATEFUNCTION'>, <TokenType.DECIMAL: 'DECIMAL'>, <TokenType.INT4MULTIRANGE: 'INT4MULTIRANGE'>, <TokenType.BIT: 'BIT'>, <TokenType.OBJECT_IDENTIFIER: 'OBJECT_IDENTIFIER'>, <TokenType.DATE32: 'DATE32'>, <TokenType.CURRENT_TIME: 'CURRENT_TIME'>, <TokenType.TIME: 'TIME'>, <TokenType.CURRENT_TIMESTAMP: 'CURRENT_TIMESTAMP'>, <TokenType.SMALLSERIAL: 'SMALLSERIAL'>, <TokenType.COMMAND: 'COMMAND'>, <TokenType.TINYBLOB: 'TINYBLOB'>, <TokenType.SUPER: 'SUPER'>, <TokenType.PERCENT: 'PERCENT'>, <TokenType.USMALLINT: 'USMALLINT'>, <TokenType.TRUNCATE: 'TRUNCATE'>, <TokenType.NCHAR: 'NCHAR'>, <TokenType.HLLSKETCH: 'HLLSKETCH'>, <TokenType.CONSTRAINT: 'CONSTRAINT'>, <TokenType.CACHE: 'CACHE'>, <TokenType.BIGSERIAL: 'BIGSERIAL'>, <TokenType.VIEW: 'VIEW'>, <TokenType.FALSE: 'FALSE'>, <TokenType.JSONB: 'JSONB'>, <TokenType.FIRST: 'FIRST'>, <TokenType.NULL: 'NULL'>, <TokenType.UINT: 'UINT'>, <TokenType.INET: 'INET'>, <TokenType.UNIQUE: 'UNIQUE'>, <TokenType.SCHEMA: 'SCHEMA'>, <TokenType.DESCRIBE: 'DESCRIBE'>, <TokenType.FILTER: 'FILTER'>, <TokenType.NULLABLE: 'NULLABLE'>, <TokenType.DATETIME: 'DATETIME'>, <TokenType.VARIANT: 'VARIANT'>, <TokenType.USE: 'USE'>, <TokenType.FIXEDSTRING: 'FIXEDSTRING'>, <TokenType.INT: 'INT'>, <TokenType.IPV6: 'IPV6'>, <TokenType.FLOAT: 'FLOAT'>, <TokenType.STORAGE_INTEGRATION: 'STORAGE_INTEGRATION'>, <TokenType.CASE: 'CASE'>, <TokenType.INT8RANGE: 'INT8RANGE'>, <TokenType.DESC: 'DESC'>, <TokenType.CURRENT_DATETIME: 'CURRENT_DATETIME'>, <TokenType.BEGIN: 'BEGIN'>, <TokenType.INT8MULTIRANGE: 'INT8MULTIRANGE'>, <TokenType.KILL: 'KILL'>, <TokenType.TSRANGE: 'TSRANGE'>, <TokenType.IPPREFIX: 'IPPREFIX'>, <TokenType.TINYINT: 'TINYINT'>, <TokenType.STRUCT: 'STRUCT'>, <TokenType.OPERATOR: 'OPERATOR'>, <TokenType.FINAL: 'FINAL'>, <TokenType.REPLACE: 'REPLACE'>, <TokenType.ENUM8: 'ENUM8'>, <TokenType.INT4RANGE: 'INT4RANGE'>, <TokenType.BOOLEAN: 'BOOLEAN'>, <TokenType.DELETE: 'DELETE'>, <TokenType.VAR: 'VAR'>, <TokenType.NVARCHAR: 'NVARCHAR'>, <TokenType.TEXT: 'TEXT'>, <TokenType.TIMESTAMP_NS: 'TIMESTAMP_NS'>, <TokenType.UPDATE: 'UPDATE'>, <TokenType.BIGDECIMAL: 'BIGDECIMAL'>, <TokenType.ROW: 'ROW'>, <TokenType.NAME: 'NAME'>, <TokenType.UINT128: 'UINT128'>, <TokenType.UNIQUEIDENTIFIER: 'UNIQUEIDENTIFIER'>, <TokenType.TRUE: 'TRUE'>, <TokenType.PRAGMA: 'PRAGMA'>, <TokenType.WAREHOUSE: 'WAREHOUSE'>, <TokenType.NUMMULTIRANGE: 'NUMMULTIRANGE'>, <TokenType.JSON: 'JSON'>, <TokenType.OVERLAPS: 'OVERLAPS'>, <TokenType.ALL: 'ALL'>, <TokenType.TIMETZ: 'TIMETZ'>, <TokenType.MEDIUMTEXT: 'MEDIUMTEXT'>, <TokenType.UDECIMAL: 'UDECIMAL'>, <TokenType.ROWS: 'ROWS'>, <TokenType.XML: 'XML'>, <TokenType.UBIGINT: 'UBIGINT'>, <TokenType.PSEUDO_TYPE: 'PSEUDO_TYPE'>, <TokenType.NEXT: 'NEXT'>, <TokenType.UNKNOWN: 'UNKNOWN'>, <TokenType.ROWVERSION: 'ROWVERSION'>, <TokenType.ANY: 'ANY'>, <TokenType.MONEY: 'MONEY'>, <TokenType.TIMESTAMP_MS: 'TIMESTAMP_MS'>, <TokenType.OVERWRITE: 'OVERWRITE'>, <TokenType.OBJECT: 'OBJECT'>, <TokenType.LONGTEXT: 'LONGTEXT'>, <TokenType.UINT256: 'UINT256'>, <TokenType.REFRESH: 'REFRESH'>, <TokenType.LIST: 'LIST'>, <TokenType.STREAMLIT: 'STREAMLIT'>, <TokenType.VARCHAR: 'VARCHAR'>, <TokenType.INT256: 'INT256'>, <TokenType.PIVOT: 'PIVOT'>, <TokenType.TIMESTAMPTZ: 'TIMESTAMPTZ'>, <TokenType.TOP: 'TOP'>, <TokenType.LOWCARDINALITY: 'LOWCARDINALITY'>, <TokenType.CURRENT_DATE: 'CURRENT_DATE'>, <TokenType.SEQUENCE: 'SEQUENCE'>, <TokenType.PROCEDURE: 'PROCEDURE'>, <TokenType.FOREIGN_KEY: 'FOREIGN_KEY'>, <TokenType.ENUM: 'ENUM'>, <TokenType.IDENTIFIER: 'IDENTIFIER'>, <TokenType.MEDIUMINT: 'MEDIUMINT'>, <TokenType.IPADDRESS: 'IPADDRESS'>, <TokenType.ANTI: 'ANTI'>, <TokenType.COPY: 'COPY'>, <TokenType.DATEMULTIRANGE: 'DATEMULTIRANGE'>, <TokenType.VARBINARY: 'VARBINARY'>, <TokenType.YEAR: 'YEAR'>, <TokenType.TSTZRANGE: 'TSTZRANGE'>, <TokenType.DIV: 'DIV'>, <TokenType.FUNCTION: 'FUNCTION'>, <TokenType.COLLATE: 'COLLATE'>, <TokenType.ENUM16: 'ENUM16'>, <TokenType.RANGE: 'RANGE'>, <TokenType.INT128: 'INT128'>, <TokenType.UTINYINT: 'UTINYINT'>, <TokenType.AUTO_INCREMENT: 'AUTO_INCREMENT'>, <TokenType.ASC: 'ASC'>, <TokenType.ROLLUP: 'ROLLUP'>, <TokenType.DATERANGE: 'DATERANGE'>, <TokenType.EXECUTE: 'EXECUTE'>, <TokenType.DATABASE: 'DATABASE'>, <TokenType.REFERENCES: 'REFERENCES'>, <TokenType.TSTZMULTIRANGE: 'TSTZMULTIRANGE'>, <TokenType.TIMESTAMPNTZ: 'TIMESTAMPNTZ'>, <TokenType.UNNEST: 'UNNEST'>, <TokenType.DOUBLE: 'DOUBLE'>, <TokenType.TINYTEXT: 'TINYTEXT'>, <TokenType.MERGE: 'MERGE'>, <TokenType.TIMESTAMP_S: 'TIMESTAMP_S'>, <TokenType.USERDEFINED: 'USERDEFINED'>, <TokenType.SEMI: 'SEMI'>, <TokenType.IMAGE: 'IMAGE'>, <TokenType.COMMENT: 'COMMENT'>, <TokenType.KEEP: 'KEEP'>, <TokenType.HSTORE: 'HSTORE'>, <TokenType.AGGREGATEFUNCTION: 'AGGREGATEFUNCTION'>, <TokenType.TAG: 'TAG'>, <TokenType.SOME: 'SOME'>, <TokenType.LONGBLOB: 'LONGBLOB'>, <TokenType.MODEL: 'MODEL'>, <TokenType.GEOMETRY: 'GEOMETRY'>, <TokenType.BIGINT: 'BIGINT'>, <TokenType.TIMESTAMP: 'TIMESTAMP'>, <TokenType.SMALLMONEY: 'SMALLMONEY'>, <TokenType.IPV4: 'IPV4'>, <TokenType.RECURSIVE: 'RECURSIVE'>, <TokenType.PARTITION: 'PARTITION'>, <TokenType.BINARY: 'BINARY'>, <TokenType.ISNULL: 'ISNULL'>, <TokenType.SHOW: 'SHOW'>, <TokenType.UNPIVOT: 'UNPIVOT'>, <TokenType.SET: 'SET'>, <TokenType.STRAIGHT_JOIN: 'STRAIGHT_JOIN'>, <TokenType.BPCHAR: 'BPCHAR'>, <TokenType.NUMRANGE: 'NUMRANGE'>, <TokenType.MEDIUMBLOB: 'MEDIUMBLOB'>, <TokenType.COLUMN: 'COLUMN'>, <TokenType.VOLATILE: 'VOLATILE'>, <TokenType.END: 'END'>, <TokenType.CHAR: 'CHAR'>, <TokenType.FORMAT: 'FORMAT'>, <TokenType.ORDINALITY: 'ORDINALITY'>, <TokenType.INDEX: 'INDEX'>, <TokenType.INTERVAL: 'INTERVAL'>, <TokenType.ARRAY: 'ARRAY'>, <TokenType.UUID: 'UUID'>, <TokenType.DEFAULT: 'DEFAULT'>, <TokenType.TIMESTAMPLTZ: 'TIMESTAMPLTZ'>, <TokenType.DATETIME64: 'DATETIME64'>, <TokenType.IS: 'IS'>, <TokenType.TEMPORARY: 'TEMPORARY'>, <TokenType.DATE: 'DATE'>, <TokenType.SETTINGS: 'SETTINGS'>, <TokenType.UMEDIUMINT: 'UMEDIUMINT'>}
SHOW_TRIE: Dict = {}
SET_TRIE: Dict = {'GLOBAL': {0: True}, 'LOCAL': {0: True}, 'SESSION': {0: True}, 'TRANSACTION': {0: True}}
Inherited Members
sqlglot.parser.Parser
Parser
STRUCT_TYPE_TOKENS
ENUM_TYPE_TOKENS
AGGREGATE_TYPE_TOKENS
TYPE_TOKENS
SIGNED_TO_UNSIGNED_TYPE_TOKEN
SUBQUERY_PREDICATES
RESERVED_TOKENS
DB_CREATABLES
CREATABLES
INTERVAL_VARS
ALIAS_TOKENS
ARRAY_CONSTRUCTORS
COMMENT_TABLE_ALIAS_TOKENS
UPDATE_ALIAS_TOKENS
TRIM_TYPES
FUNC_TOKENS
CONJUNCTION
ASSIGNMENT
DISJUNCTION
EQUALITY
COMPARISON
BITWISE
TERM
FACTOR
EXPONENT
TIMES
TIMESTAMPS
SET_OPERATIONS
JOIN_METHODS
JOIN_SIDES
JOIN_KINDS
JOIN_HINTS
LAMBDAS
COLUMN_OPERATORS
EXPRESSION_PARSERS
UNARY_PARSERS
STRING_PARSERS
NUMERIC_PARSERS
PRIMARY_PARSERS
PLACEHOLDER_PARSERS
ALTER_PARSERS
ALTER_ALTER_PARSERS
SCHEMA_UNNAMED_CONSTRAINTS
NO_PAREN_FUNCTION_PARSERS
INVALID_FUNC_NAME_TOKENS
FUNCTIONS_WITH_ALIASED_ARGS
KEY_VALUE_DEFINITIONS
QUERY_MODIFIER_PARSERS
SET_PARSERS
SHOW_PARSERS
TYPE_LITERAL_PARSERS
TYPE_CONVERTERS
DDL_SELECT_TOKENS
PRE_VOLATILE_TOKENS
TRANSACTION_KIND
TRANSACTION_CHARACTERISTICS
CONFLICT_ACTIONS
CREATE_SEQUENCE
ISOLATED_LOADING_OPTIONS
USABLES
CAST_ACTIONS
INSERT_ALTERNATIVES
CLONE_KEYWORDS
HISTORICAL_DATA_KIND
OPCLASS_FOLLOW_KEYWORDS
OPTYPE_FOLLOW_TOKENS
TABLE_INDEX_HINT_TOKENS
VIEW_ATTRIBUTES
WINDOW_ALIAS_TOKENS
WINDOW_BEFORE_PAREN_TOKENS
WINDOW_SIDES
JSON_KEY_VALUE_SEPARATOR_TOKENS
FETCH_TOKENS
ADD_CONSTRAINT_TOKENS
DISTINCT_TOKENS
UNNEST_OFFSET_ALIAS_TOKENS
SELECT_START_TOKENS
COPY_INTO_VARLEN_OPTIONS
STRICT_CAST
IDENTIFY_PIVOT_STRINGS
ALTER_TABLE_ADD_REQUIRED_FOR_EACH_COLUMN
TABLESAMPLE_CSV
DEFAULT_SAMPLING_METHOD
SET_REQUIRES_ASSIGNMENT_DELIMITER
TRIM_PATTERN_FIRST
STRING_ALIASES
MODIFIERS_ATTACHED_TO_SET_OP
SET_OP_MODIFIERS
NO_PAREN_IF_COMMANDS
JSON_ARROWS_REQUIRE_JSON_TYPE
COLON_IS_JSON_EXTRACT
VALUES_FOLLOWED_BY_PAREN
INTERVAL_SPANS
SUPPORTS_PARTITION_SELECTION
error_level
error_message_context
max_errors
dialect
reset
parse
parse_into
check_errors
raise_error
expression
validate_expression
errors
sql
class BigQuery.Generator(sqlglot.generator.Generator):
564    class Generator(generator.Generator):
565        EXPLICIT_SET_OP = True
566        INTERVAL_ALLOWS_PLURAL_FORM = False
567        JOIN_HINTS = False
568        QUERY_HINTS = False
569        TABLE_HINTS = False
570        LIMIT_FETCH = "LIMIT"
571        RENAME_TABLE_WITH_DB = False
572        NVL2_SUPPORTED = False
573        UNNEST_WITH_ORDINALITY = False
574        COLLATE_IS_FUNC = True
575        LIMIT_ONLY_LITERALS = True
576        SUPPORTS_TABLE_ALIAS_COLUMNS = False
577        UNPIVOT_ALIASES_ARE_IDENTIFIERS = False
578        JSON_KEY_VALUE_PAIR_SEP = ","
579        NULL_ORDERING_SUPPORTED = False
580        IGNORE_NULLS_IN_FUNC = True
581        JSON_PATH_SINGLE_QUOTE_ESCAPE = True
582        CAN_IMPLEMENT_ARRAY_ANY = True
583        SUPPORTS_TO_NUMBER = False
584        NAMED_PLACEHOLDER_TOKEN = "@"
585        HEX_FUNC = "TO_HEX"
586        WITH_PROPERTIES_PREFIX = "OPTIONS"
587
588        TRANSFORMS = {
589            **generator.Generator.TRANSFORMS,
590            exp.ApproxDistinct: rename_func("APPROX_COUNT_DISTINCT"),
591            exp.ArgMax: arg_max_or_min_no_count("MAX_BY"),
592            exp.ArgMin: arg_max_or_min_no_count("MIN_BY"),
593            exp.Array: inline_array_unless_query,
594            exp.ArrayContains: _array_contains_sql,
595            exp.ArrayFilter: filter_array_using_unnest,
596            exp.ArraySize: rename_func("ARRAY_LENGTH"),
597            exp.Cast: transforms.preprocess([transforms.remove_precision_parameterized_types]),
598            exp.CollateProperty: lambda self, e: (
599                f"DEFAULT COLLATE {self.sql(e, 'this')}"
600                if e.args.get("default")
601                else f"COLLATE {self.sql(e, 'this')}"
602            ),
603            exp.Commit: lambda *_: "COMMIT TRANSACTION",
604            exp.CountIf: rename_func("COUNTIF"),
605            exp.Create: _create_sql,
606            exp.CTE: transforms.preprocess([_pushdown_cte_column_names]),
607            exp.DateAdd: date_add_interval_sql("DATE", "ADD"),
608            exp.DateDiff: lambda self, e: self.func(
609                "DATE_DIFF", e.this, e.expression, unit_to_var(e)
610            ),
611            exp.DateFromParts: rename_func("DATE"),
612            exp.DateStrToDate: datestrtodate_sql,
613            exp.DateSub: date_add_interval_sql("DATE", "SUB"),
614            exp.DatetimeAdd: date_add_interval_sql("DATETIME", "ADD"),
615            exp.DatetimeSub: date_add_interval_sql("DATETIME", "SUB"),
616            exp.DateTrunc: lambda self, e: self.func("DATE_TRUNC", e.this, e.text("unit")),
617            exp.FromTimeZone: lambda self, e: self.func(
618                "DATETIME", self.func("TIMESTAMP", e.this, e.args.get("zone")), "'UTC'"
619            ),
620            exp.GenerateSeries: rename_func("GENERATE_ARRAY"),
621            exp.GroupConcat: rename_func("STRING_AGG"),
622            exp.Hex: lambda self, e: self.func("UPPER", self.func("TO_HEX", self.sql(e, "this"))),
623            exp.If: if_sql(false_value="NULL"),
624            exp.ILike: no_ilike_sql,
625            exp.IntDiv: rename_func("DIV"),
626            exp.JSONFormat: rename_func("TO_JSON_STRING"),
627            exp.Max: max_or_greatest,
628            exp.MD5: lambda self, e: self.func("TO_HEX", self.func("MD5", e.this)),
629            exp.MD5Digest: rename_func("MD5"),
630            exp.Min: min_or_least,
631            exp.PartitionedByProperty: lambda self, e: f"PARTITION BY {self.sql(e, 'this')}",
632            exp.RegexpExtract: lambda self, e: self.func(
633                "REGEXP_EXTRACT",
634                e.this,
635                e.expression,
636                e.args.get("position"),
637                e.args.get("occurrence"),
638            ),
639            exp.RegexpReplace: regexp_replace_sql,
640            exp.RegexpLike: rename_func("REGEXP_CONTAINS"),
641            exp.ReturnsProperty: _returnsproperty_sql,
642            exp.Rollback: lambda *_: "ROLLBACK TRANSACTION",
643            exp.Select: transforms.preprocess(
644                [
645                    transforms.explode_to_unnest(),
646                    transforms.unqualify_unnest,
647                    transforms.eliminate_distinct_on,
648                    _alias_ordered_group,
649                    transforms.eliminate_semi_and_anti_joins,
650                ]
651            ),
652            exp.SHA: rename_func("SHA1"),
653            exp.SHA2: sha256_sql,
654            exp.StabilityProperty: lambda self, e: (
655                "DETERMINISTIC" if e.name == "IMMUTABLE" else "NOT DETERMINISTIC"
656            ),
657            exp.StrToDate: lambda self, e: self.func("PARSE_DATE", self.format_time(e), e.this),
658            exp.StrToTime: lambda self, e: self.func(
659                "PARSE_TIMESTAMP", self.format_time(e), e.this, e.args.get("zone")
660            ),
661            exp.TimeAdd: date_add_interval_sql("TIME", "ADD"),
662            exp.TimeFromParts: rename_func("TIME"),
663            exp.TimestampFromParts: rename_func("DATETIME"),
664            exp.TimeSub: date_add_interval_sql("TIME", "SUB"),
665            exp.TimestampAdd: date_add_interval_sql("TIMESTAMP", "ADD"),
666            exp.TimestampDiff: rename_func("TIMESTAMP_DIFF"),
667            exp.TimestampSub: date_add_interval_sql("TIMESTAMP", "SUB"),
668            exp.TimeStrToTime: timestrtotime_sql,
669            exp.Transaction: lambda *_: "BEGIN TRANSACTION",
670            exp.Trim: lambda self, e: self.func("TRIM", e.this, e.expression),
671            exp.TsOrDsAdd: _ts_or_ds_add_sql,
672            exp.TsOrDsDiff: _ts_or_ds_diff_sql,
673            exp.TsOrDsToTime: rename_func("TIME"),
674            exp.TsOrDsToTimestamp: rename_func("DATETIME"),
675            exp.Unhex: rename_func("FROM_HEX"),
676            exp.UnixDate: rename_func("UNIX_DATE"),
677            exp.UnixToTime: _unix_to_time_sql,
678            exp.Values: _derived_table_values_to_unnest,
679            exp.VariancePop: rename_func("VAR_POP"),
680        }
681
682        SUPPORTED_JSON_PATH_PARTS = {
683            exp.JSONPathKey,
684            exp.JSONPathRoot,
685            exp.JSONPathSubscript,
686        }
687
688        TYPE_MAPPING = {
689            **generator.Generator.TYPE_MAPPING,
690            exp.DataType.Type.BIGDECIMAL: "BIGNUMERIC",
691            exp.DataType.Type.BIGINT: "INT64",
692            exp.DataType.Type.BINARY: "BYTES",
693            exp.DataType.Type.BOOLEAN: "BOOL",
694            exp.DataType.Type.CHAR: "STRING",
695            exp.DataType.Type.DECIMAL: "NUMERIC",
696            exp.DataType.Type.DOUBLE: "FLOAT64",
697            exp.DataType.Type.FLOAT: "FLOAT64",
698            exp.DataType.Type.INT: "INT64",
699            exp.DataType.Type.NCHAR: "STRING",
700            exp.DataType.Type.NVARCHAR: "STRING",
701            exp.DataType.Type.SMALLINT: "INT64",
702            exp.DataType.Type.TEXT: "STRING",
703            exp.DataType.Type.TIMESTAMP: "DATETIME",
704            exp.DataType.Type.TIMESTAMPTZ: "TIMESTAMP",
705            exp.DataType.Type.TIMESTAMPLTZ: "TIMESTAMP",
706            exp.DataType.Type.TINYINT: "INT64",
707            exp.DataType.Type.VARBINARY: "BYTES",
708            exp.DataType.Type.ROWVERSION: "BYTES",
709            exp.DataType.Type.VARCHAR: "STRING",
710            exp.DataType.Type.VARIANT: "ANY TYPE",
711        }
712
713        PROPERTIES_LOCATION = {
714            **generator.Generator.PROPERTIES_LOCATION,
715            exp.PartitionedByProperty: exp.Properties.Location.POST_SCHEMA,
716            exp.VolatileProperty: exp.Properties.Location.UNSUPPORTED,
717        }
718
719        # from: https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#reserved_keywords
720        RESERVED_KEYWORDS = {
721            "all",
722            "and",
723            "any",
724            "array",
725            "as",
726            "asc",
727            "assert_rows_modified",
728            "at",
729            "between",
730            "by",
731            "case",
732            "cast",
733            "collate",
734            "contains",
735            "create",
736            "cross",
737            "cube",
738            "current",
739            "default",
740            "define",
741            "desc",
742            "distinct",
743            "else",
744            "end",
745            "enum",
746            "escape",
747            "except",
748            "exclude",
749            "exists",
750            "extract",
751            "false",
752            "fetch",
753            "following",
754            "for",
755            "from",
756            "full",
757            "group",
758            "grouping",
759            "groups",
760            "hash",
761            "having",
762            "if",
763            "ignore",
764            "in",
765            "inner",
766            "intersect",
767            "interval",
768            "into",
769            "is",
770            "join",
771            "lateral",
772            "left",
773            "like",
774            "limit",
775            "lookup",
776            "merge",
777            "natural",
778            "new",
779            "no",
780            "not",
781            "null",
782            "nulls",
783            "of",
784            "on",
785            "or",
786            "order",
787            "outer",
788            "over",
789            "partition",
790            "preceding",
791            "proto",
792            "qualify",
793            "range",
794            "recursive",
795            "respect",
796            "right",
797            "rollup",
798            "rows",
799            "select",
800            "set",
801            "some",
802            "struct",
803            "tablesample",
804            "then",
805            "to",
806            "treat",
807            "true",
808            "unbounded",
809            "union",
810            "unnest",
811            "using",
812            "when",
813            "where",
814            "window",
815            "with",
816            "within",
817        }
818
819        def mod_sql(self, expression: exp.Mod) -> str:
820            this = expression.this
821            expr = expression.expression
822            return self.func(
823                "MOD",
824                this.unnest() if isinstance(this, exp.Paren) else this,
825                expr.unnest() if isinstance(expr, exp.Paren) else expr,
826            )
827
828        def column_parts(self, expression: exp.Column) -> str:
829            if expression.meta.get("quoted_column"):
830                # If a column reference is of the form `dataset.table`.name, we need
831                # to preserve the quoted table path, otherwise the reference breaks
832                table_parts = ".".join(p.name for p in expression.parts[:-1])
833                table_path = self.sql(exp.Identifier(this=table_parts, quoted=True))
834                return f"{table_path}.{self.sql(expression, 'this')}"
835
836            return super().column_parts(expression)
837
838        def table_parts(self, expression: exp.Table) -> str:
839            # Depending on the context, `x.y` may not resolve to the same data source as `x`.`y`, so
840            # we need to make sure the correct quoting is used in each case.
841            #
842            # For example, if there is a CTE x that clashes with a schema name, then the former will
843            # return the table y in that schema, whereas the latter will return the CTE's y column:
844            #
845            # - WITH x AS (SELECT [1, 2] AS y) SELECT * FROM x, `x.y`   -> cross join
846            # - WITH x AS (SELECT [1, 2] AS y) SELECT * FROM x, `x`.`y` -> implicit unnest
847            if expression.meta.get("quoted_table"):
848                table_parts = ".".join(p.name for p in expression.parts)
849                return self.sql(exp.Identifier(this=table_parts, quoted=True))
850
851            return super().table_parts(expression)
852
853        def timetostr_sql(self, expression: exp.TimeToStr) -> str:
854            this = expression.this if isinstance(expression.this, exp.TsOrDsToDate) else expression
855            return self.func("FORMAT_DATE", self.format_time(expression), this.this)
856
857        def eq_sql(self, expression: exp.EQ) -> str:
858            # Operands of = cannot be NULL in BigQuery
859            if isinstance(expression.left, exp.Null) or isinstance(expression.right, exp.Null):
860                if not isinstance(expression.parent, exp.Update):
861                    return "NULL"
862
863            return self.binary(expression, "=")
864
865        def attimezone_sql(self, expression: exp.AtTimeZone) -> str:
866            parent = expression.parent
867
868            # BigQuery allows CAST(.. AS {STRING|TIMESTAMP} [FORMAT <fmt> [AT TIME ZONE <tz>]]).
869            # Only the TIMESTAMP one should use the below conversion, when AT TIME ZONE is included.
870            if not isinstance(parent, exp.Cast) or not parent.to.is_type("text"):
871                return self.func(
872                    "TIMESTAMP", self.func("DATETIME", expression.this, expression.args.get("zone"))
873                )
874
875            return super().attimezone_sql(expression)
876
877        def trycast_sql(self, expression: exp.TryCast) -> str:
878            return self.cast_sql(expression, safe_prefix="SAFE_")
879
880        def bracket_sql(self, expression: exp.Bracket) -> str:
881            this = expression.this
882            expressions = expression.expressions
883
884            if len(expressions) == 1 and this and this.is_type(exp.DataType.Type.STRUCT):
885                arg = expressions[0]
886                if arg.type is None:
887                    from sqlglot.optimizer.annotate_types import annotate_types
888
889                    arg = annotate_types(arg)
890
891                if arg.type and arg.type.this in exp.DataType.TEXT_TYPES:
892                    # BQ doesn't support bracket syntax with string values for structs
893                    return f"{self.sql(this)}.{arg.name}"
894
895            expressions_sql = self.expressions(expression, flat=True)
896            offset = expression.args.get("offset")
897
898            if offset == 0:
899                expressions_sql = f"OFFSET({expressions_sql})"
900            elif offset == 1:
901                expressions_sql = f"ORDINAL({expressions_sql})"
902            elif offset is not None:
903                self.unsupported(f"Unsupported array offset: {offset}")
904
905            if expression.args.get("safe"):
906                expressions_sql = f"SAFE_{expressions_sql}"
907
908            return f"{self.sql(this)}[{expressions_sql}]"
909
910        def in_unnest_op(self, expression: exp.Unnest) -> str:
911            return self.sql(expression)
912
913        def except_op(self, expression: exp.Except) -> str:
914            if not expression.args.get("distinct"):
915                self.unsupported("EXCEPT without DISTINCT is not supported in BigQuery")
916            return f"EXCEPT{' DISTINCT' if expression.args.get('distinct') else ' ALL'}"
917
918        def intersect_op(self, expression: exp.Intersect) -> str:
919            if not expression.args.get("distinct"):
920                self.unsupported("INTERSECT without DISTINCT is not supported in BigQuery")
921            return f"INTERSECT{' DISTINCT' if expression.args.get('distinct') else ' ALL'}"
922
923        def version_sql(self, expression: exp.Version) -> str:
924            if expression.name == "TIMESTAMP":
925                expression.set("this", "SYSTEM_TIME")
926            return super().version_sql(expression)

Generator converts a given syntax tree to the corresponding SQL string.

Arguments:
  • pretty: Whether to format the produced SQL string. Default: False.
  • identify: Determines when an identifier should be quoted. Possible values are: False (default): Never quote, except in cases where it's mandatory by the dialect. True or 'always': Always quote. 'safe': Only quote identifiers that are case insensitive.
  • normalize: Whether to normalize identifiers to lowercase. Default: False.
  • pad: The pad size in a formatted string. For example, this affects the indentation of a projection in a query, relative to its nesting level. Default: 2.
  • indent: The indentation size in a formatted string. For example, this affects the indentation of subqueries and filters under a WHERE clause. Default: 2.
  • normalize_functions: How to normalize function names. Possible values are: "upper" or True (default): Convert names to uppercase. "lower": Convert names to lowercase. False: Disables function name normalization.
  • unsupported_level: Determines the generator's behavior when it encounters unsupported expressions. Default ErrorLevel.WARN.
  • max_unsupported: Maximum number of unsupported messages to include in a raised UnsupportedError. This is only relevant if unsupported_level is ErrorLevel.RAISE. Default: 3
  • leading_comma: Whether the comma is leading or trailing in select expressions. This is only relevant when generating in pretty mode. Default: False
  • max_text_width: The max number of characters in a segment before creating new lines in pretty mode. The default is on the smaller end because the length only represents a segment and not the true line length. Default: 80
  • comments: Whether to preserve comments in the output SQL code. Default: True
EXPLICIT_SET_OP = True
INTERVAL_ALLOWS_PLURAL_FORM = False
JOIN_HINTS = False
QUERY_HINTS = False
TABLE_HINTS = False
LIMIT_FETCH = 'LIMIT'
RENAME_TABLE_WITH_DB = False
NVL2_SUPPORTED = False
UNNEST_WITH_ORDINALITY = False
COLLATE_IS_FUNC = True
LIMIT_ONLY_LITERALS = True
SUPPORTS_TABLE_ALIAS_COLUMNS = False
UNPIVOT_ALIASES_ARE_IDENTIFIERS = False
JSON_KEY_VALUE_PAIR_SEP = ','
NULL_ORDERING_SUPPORTED = False
IGNORE_NULLS_IN_FUNC = True
JSON_PATH_SINGLE_QUOTE_ESCAPE = True
CAN_IMPLEMENT_ARRAY_ANY = True
SUPPORTS_TO_NUMBER = False
NAMED_PLACEHOLDER_TOKEN = '@'
HEX_FUNC = 'TO_HEX'
WITH_PROPERTIES_PREFIX = 'OPTIONS'
TRANSFORMS = {<class 'sqlglot.expressions.JSONPathKey'>: <function <lambda>>, <class 'sqlglot.expressions.JSONPathRoot'>: <function <lambda>>, <class 'sqlglot.expressions.JSONPathSubscript'>: <function <lambda>>, <class 'sqlglot.expressions.AllowedValuesProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.AutoRefreshProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.BackupProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.CaseSpecificColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.CharacterSetColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.CharacterSetProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ClusteredColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.CollateColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.CommentColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.CopyGrantsProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.DateFormatColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.DefaultColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.EncodeColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.EphemeralColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ExcludeColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ExecuteAsProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ExternalProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.GlobalProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.HeapProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.IcebergProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.InheritsProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.InlineLengthColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.InputModelProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.IntervalSpan'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.JSONExtract'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.JSONExtractScalar'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.LanguageProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.LocationProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.LogProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.MaterializedProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.NonClusteredColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.NoPrimaryIndexProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.NotForReplicationColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.OnCommitProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.OnProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.OnUpdateColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.OutputModelProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.PathColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ProjectionPolicyColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.RemoteWithConnectionModelProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ReturnsProperty'>: <function _returnsproperty_sql>, <class 'sqlglot.expressions.SampleProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SecureProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SetConfigProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SetProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SettingsProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SharingProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SqlReadWriteProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SqlSecurityProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.StabilityProperty'>: <function BigQuery.Generator.<lambda>>, <class 'sqlglot.expressions.StrictProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.TemporaryProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.TagColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.TitleColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.Timestamp'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ToMap'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ToTableProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.TransformModelProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.TransientProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.UppercaseColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.UnloggedProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.VarMap'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ViewAttributeProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.VolatileProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.WithJournalTableProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.WithOperator'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ApproxDistinct'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.ArgMax'>: <function arg_max_or_min_no_count.<locals>._arg_max_or_min_sql>, <class 'sqlglot.expressions.ArgMin'>: <function arg_max_or_min_no_count.<locals>._arg_max_or_min_sql>, <class 'sqlglot.expressions.Array'>: <function inline_array_unless_query>, <class 'sqlglot.expressions.ArrayContains'>: <function _array_contains_sql>, <class 'sqlglot.expressions.ArrayFilter'>: <function filter_array_using_unnest>, <class 'sqlglot.expressions.ArraySize'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.Cast'>: <function preprocess.<locals>._to_sql>, <class 'sqlglot.expressions.CollateProperty'>: <function BigQuery.Generator.<lambda>>, <class 'sqlglot.expressions.Commit'>: <function BigQuery.Generator.<lambda>>, <class 'sqlglot.expressions.CountIf'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.Create'>: <function _create_sql>, <class 'sqlglot.expressions.CTE'>: <function preprocess.<locals>._to_sql>, <class 'sqlglot.expressions.DateAdd'>: <function date_add_interval_sql.<locals>.func>, <class 'sqlglot.expressions.DateDiff'>: <function BigQuery.Generator.<lambda>>, <class 'sqlglot.expressions.DateFromParts'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.DateStrToDate'>: <function datestrtodate_sql>, <class 'sqlglot.expressions.DateSub'>: <function date_add_interval_sql.<locals>.func>, <class 'sqlglot.expressions.DatetimeAdd'>: <function date_add_interval_sql.<locals>.func>, <class 'sqlglot.expressions.DatetimeSub'>: <function date_add_interval_sql.<locals>.func>, <class 'sqlglot.expressions.DateTrunc'>: <function BigQuery.Generator.<lambda>>, <class 'sqlglot.expressions.FromTimeZone'>: <function BigQuery.Generator.<lambda>>, <class 'sqlglot.expressions.GenerateSeries'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.GroupConcat'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.Hex'>: <function BigQuery.Generator.<lambda>>, <class 'sqlglot.expressions.If'>: <function if_sql.<locals>._if_sql>, <class 'sqlglot.expressions.ILike'>: <function no_ilike_sql>, <class 'sqlglot.expressions.IntDiv'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.JSONFormat'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.Max'>: <function max_or_greatest>, <class 'sqlglot.expressions.MD5'>: <function BigQuery.Generator.<lambda>>, <class 'sqlglot.expressions.MD5Digest'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.Min'>: <function min_or_least>, <class 'sqlglot.expressions.PartitionedByProperty'>: <function BigQuery.Generator.<lambda>>, <class 'sqlglot.expressions.RegexpExtract'>: <function BigQuery.Generator.<lambda>>, <class 'sqlglot.expressions.RegexpReplace'>: <function regexp_replace_sql>, <class 'sqlglot.expressions.RegexpLike'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.Rollback'>: <function BigQuery.Generator.<lambda>>, <class 'sqlglot.expressions.Select'>: <function preprocess.<locals>._to_sql>, <class 'sqlglot.expressions.SHA'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.SHA2'>: <function sha256_sql>, <class 'sqlglot.expressions.StrToDate'>: <function BigQuery.Generator.<lambda>>, <class 'sqlglot.expressions.StrToTime'>: <function BigQuery.Generator.<lambda>>, <class 'sqlglot.expressions.TimeAdd'>: <function date_add_interval_sql.<locals>.func>, <class 'sqlglot.expressions.TimeFromParts'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.TimestampFromParts'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.TimeSub'>: <function date_add_interval_sql.<locals>.func>, <class 'sqlglot.expressions.TimestampAdd'>: <function date_add_interval_sql.<locals>.func>, <class 'sqlglot.expressions.TimestampDiff'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.TimestampSub'>: <function date_add_interval_sql.<locals>.func>, <class 'sqlglot.expressions.TimeStrToTime'>: <function timestrtotime_sql>, <class 'sqlglot.expressions.Transaction'>: <function BigQuery.Generator.<lambda>>, <class 'sqlglot.expressions.Trim'>: <function BigQuery.Generator.<lambda>>, <class 'sqlglot.expressions.TsOrDsAdd'>: <function _ts_or_ds_add_sql>, <class 'sqlglot.expressions.TsOrDsDiff'>: <function _ts_or_ds_diff_sql>, <class 'sqlglot.expressions.TsOrDsToTime'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.TsOrDsToTimestamp'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.Unhex'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.UnixDate'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.UnixToTime'>: <function _unix_to_time_sql>, <class 'sqlglot.expressions.Values'>: <function _derived_table_values_to_unnest>, <class 'sqlglot.expressions.VariancePop'>: <function rename_func.<locals>.<lambda>>}
TYPE_MAPPING = {<Type.NCHAR: 'NCHAR'>: 'STRING', <Type.NVARCHAR: 'NVARCHAR'>: 'STRING', <Type.MEDIUMTEXT: 'MEDIUMTEXT'>: 'TEXT', <Type.LONGTEXT: 'LONGTEXT'>: 'TEXT', <Type.TINYTEXT: 'TINYTEXT'>: 'TEXT', <Type.MEDIUMBLOB: 'MEDIUMBLOB'>: 'BLOB', <Type.LONGBLOB: 'LONGBLOB'>: 'BLOB', <Type.TINYBLOB: 'TINYBLOB'>: 'BLOB', <Type.INET: 'INET'>: 'INET', <Type.ROWVERSION: 'ROWVERSION'>: 'BYTES', <Type.BIGDECIMAL: 'BIGDECIMAL'>: 'BIGNUMERIC', <Type.BIGINT: 'BIGINT'>: 'INT64', <Type.BINARY: 'BINARY'>: 'BYTES', <Type.BOOLEAN: 'BOOLEAN'>: 'BOOL', <Type.CHAR: 'CHAR'>: 'STRING', <Type.DECIMAL: 'DECIMAL'>: 'NUMERIC', <Type.DOUBLE: 'DOUBLE'>: 'FLOAT64', <Type.FLOAT: 'FLOAT'>: 'FLOAT64', <Type.INT: 'INT'>: 'INT64', <Type.SMALLINT: 'SMALLINT'>: 'INT64', <Type.TEXT: 'TEXT'>: 'STRING', <Type.TIMESTAMP: 'TIMESTAMP'>: 'DATETIME', <Type.TIMESTAMPTZ: 'TIMESTAMPTZ'>: 'TIMESTAMP', <Type.TIMESTAMPLTZ: 'TIMESTAMPLTZ'>: 'TIMESTAMP', <Type.TINYINT: 'TINYINT'>: 'INT64', <Type.VARBINARY: 'VARBINARY'>: 'BYTES', <Type.VARCHAR: 'VARCHAR'>: 'STRING', <Type.VARIANT: 'VARIANT'>: 'ANY TYPE'}
PROPERTIES_LOCATION = {<class 'sqlglot.expressions.AllowedValuesProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.AlgorithmProperty'>: <Location.POST_CREATE: 'POST_CREATE'>, <class 'sqlglot.expressions.AutoIncrementProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.AutoRefreshProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.BackupProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.BlockCompressionProperty'>: <Location.POST_NAME: 'POST_NAME'>, <class 'sqlglot.expressions.CharacterSetProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.ChecksumProperty'>: <Location.POST_NAME: 'POST_NAME'>, <class 'sqlglot.expressions.CollateProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.CopyGrantsProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.Cluster'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.ClusteredByProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.DataBlocksizeProperty'>: <Location.POST_NAME: 'POST_NAME'>, <class 'sqlglot.expressions.DataDeletionProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.DefinerProperty'>: <Location.POST_CREATE: 'POST_CREATE'>, <class 'sqlglot.expressions.DictRange'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.DictProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.DistKeyProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.DistStyleProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.EngineProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.ExecuteAsProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.ExternalProperty'>: <Location.POST_CREATE: 'POST_CREATE'>, <class 'sqlglot.expressions.FallbackProperty'>: <Location.POST_NAME: 'POST_NAME'>, <class 'sqlglot.expressions.FileFormatProperty'>: <Location.POST_WITH: 'POST_WITH'>, <class 'sqlglot.expressions.FreespaceProperty'>: <Location.POST_NAME: 'POST_NAME'>, <class 'sqlglot.expressions.GlobalProperty'>: <Location.POST_CREATE: 'POST_CREATE'>, <class 'sqlglot.expressions.HeapProperty'>: <Location.POST_WITH: 'POST_WITH'>, <class 'sqlglot.expressions.InheritsProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.IcebergProperty'>: <Location.POST_CREATE: 'POST_CREATE'>, <class 'sqlglot.expressions.InputModelProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.IsolatedLoadingProperty'>: <Location.POST_NAME: 'POST_NAME'>, <class 'sqlglot.expressions.JournalProperty'>: <Location.POST_NAME: 'POST_NAME'>, <class 'sqlglot.expressions.LanguageProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.LikeProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.LocationProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.LockProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.LockingProperty'>: <Location.POST_ALIAS: 'POST_ALIAS'>, <class 'sqlglot.expressions.LogProperty'>: <Location.POST_NAME: 'POST_NAME'>, <class 'sqlglot.expressions.MaterializedProperty'>: <Location.POST_CREATE: 'POST_CREATE'>, <class 'sqlglot.expressions.MergeBlockRatioProperty'>: <Location.POST_NAME: 'POST_NAME'>, <class 'sqlglot.expressions.NoPrimaryIndexProperty'>: <Location.POST_EXPRESSION: 'POST_EXPRESSION'>, <class 'sqlglot.expressions.OnProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.OnCommitProperty'>: <Location.POST_EXPRESSION: 'POST_EXPRESSION'>, <class 'sqlglot.expressions.Order'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.OutputModelProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.PartitionedByProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.PartitionedOfProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.PrimaryKey'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.Property'>: <Location.POST_WITH: 'POST_WITH'>, <class 'sqlglot.expressions.RemoteWithConnectionModelProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.ReturnsProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.RowFormatProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.RowFormatDelimitedProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.RowFormatSerdeProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.SampleProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.SchemaCommentProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.SecureProperty'>: <Location.POST_CREATE: 'POST_CREATE'>, <class 'sqlglot.expressions.SerdeProperties'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.Set'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.SettingsProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.SetProperty'>: <Location.POST_CREATE: 'POST_CREATE'>, <class 'sqlglot.expressions.SetConfigProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.SharingProperty'>: <Location.POST_EXPRESSION: 'POST_EXPRESSION'>, <class 'sqlglot.expressions.SequenceProperties'>: <Location.POST_EXPRESSION: 'POST_EXPRESSION'>, <class 'sqlglot.expressions.SortKeyProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.SqlReadWriteProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.SqlSecurityProperty'>: <Location.POST_CREATE: 'POST_CREATE'>, <class 'sqlglot.expressions.StabilityProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.StrictProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.TemporaryProperty'>: <Location.POST_CREATE: 'POST_CREATE'>, <class 'sqlglot.expressions.ToTableProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.TransientProperty'>: <Location.POST_CREATE: 'POST_CREATE'>, <class 'sqlglot.expressions.TransformModelProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.MergeTreeTTL'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.UnloggedProperty'>: <Location.POST_CREATE: 'POST_CREATE'>, <class 'sqlglot.expressions.ViewAttributeProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.VolatileProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.WithDataProperty'>: <Location.POST_EXPRESSION: 'POST_EXPRESSION'>, <class 'sqlglot.expressions.WithJournalTableProperty'>: <Location.POST_NAME: 'POST_NAME'>, <class 'sqlglot.expressions.WithSystemVersioningProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>}
RESERVED_KEYWORDS = {'recursive', 'fetch', 'range', 'and', 'within', 'on', 'current', 'qualify', 'natural', 'treat', 'true', 'with', 'where', 'partition', 'cross', 'some', 'from', 'lateral', 'default', 'or', 'inner', 'all', 'extract', 'having', 'else', 'escape', 'no', 'cube', 'collate', 'left', 'by', 'struct', 'preceding', 'right', 'any', 'using', 'respect', 'rollup', 'select', 'if', 'rows', 'cast', 'like', 'new', 'as', 'except', 'false', 'assert_rows_modified', 'set', 'define', 'create', 'for', 'asc', 'in', 'is', 'full', 'proto', 'array', 'of', 'over', 'interval', 'enum', 'window', 'end', 'between', 'group', 'null', 'not', 'distinct', 'to', 'ignore', 'case', 'lookup', 'groups', 'limit', 'grouping', 'unnest', 'nulls', 'exclude', 'then', 'contains', 'following', 'order', 'desc', 'into', 'union', 'join', 'at', 'when', 'hash', 'exists', 'merge', 'intersect', 'tablesample', 'unbounded', 'outer'}
def mod_sql(self, expression: sqlglot.expressions.Mod) -> str:
819        def mod_sql(self, expression: exp.Mod) -> str:
820            this = expression.this
821            expr = expression.expression
822            return self.func(
823                "MOD",
824                this.unnest() if isinstance(this, exp.Paren) else this,
825                expr.unnest() if isinstance(expr, exp.Paren) else expr,
826            )
def column_parts(self, expression: sqlglot.expressions.Column) -> str:
828        def column_parts(self, expression: exp.Column) -> str:
829            if expression.meta.get("quoted_column"):
830                # If a column reference is of the form `dataset.table`.name, we need
831                # to preserve the quoted table path, otherwise the reference breaks
832                table_parts = ".".join(p.name for p in expression.parts[:-1])
833                table_path = self.sql(exp.Identifier(this=table_parts, quoted=True))
834                return f"{table_path}.{self.sql(expression, 'this')}"
835
836            return super().column_parts(expression)
def table_parts(self, expression: sqlglot.expressions.Table) -> str:
838        def table_parts(self, expression: exp.Table) -> str:
839            # Depending on the context, `x.y` may not resolve to the same data source as `x`.`y`, so
840            # we need to make sure the correct quoting is used in each case.
841            #
842            # For example, if there is a CTE x that clashes with a schema name, then the former will
843            # return the table y in that schema, whereas the latter will return the CTE's y column:
844            #
845            # - WITH x AS (SELECT [1, 2] AS y) SELECT * FROM x, `x.y`   -> cross join
846            # - WITH x AS (SELECT [1, 2] AS y) SELECT * FROM x, `x`.`y` -> implicit unnest
847            if expression.meta.get("quoted_table"):
848                table_parts = ".".join(p.name for p in expression.parts)
849                return self.sql(exp.Identifier(this=table_parts, quoted=True))
850
851            return super().table_parts(expression)
def timetostr_sql(self, expression: sqlglot.expressions.TimeToStr) -> str:
853        def timetostr_sql(self, expression: exp.TimeToStr) -> str:
854            this = expression.this if isinstance(expression.this, exp.TsOrDsToDate) else expression
855            return self.func("FORMAT_DATE", self.format_time(expression), this.this)
def eq_sql(self, expression: sqlglot.expressions.EQ) -> str:
857        def eq_sql(self, expression: exp.EQ) -> str:
858            # Operands of = cannot be NULL in BigQuery
859            if isinstance(expression.left, exp.Null) or isinstance(expression.right, exp.Null):
860                if not isinstance(expression.parent, exp.Update):
861                    return "NULL"
862
863            return self.binary(expression, "=")
def attimezone_sql(self, expression: sqlglot.expressions.AtTimeZone) -> str:
865        def attimezone_sql(self, expression: exp.AtTimeZone) -> str:
866            parent = expression.parent
867
868            # BigQuery allows CAST(.. AS {STRING|TIMESTAMP} [FORMAT <fmt> [AT TIME ZONE <tz>]]).
869            # Only the TIMESTAMP one should use the below conversion, when AT TIME ZONE is included.
870            if not isinstance(parent, exp.Cast) or not parent.to.is_type("text"):
871                return self.func(
872                    "TIMESTAMP", self.func("DATETIME", expression.this, expression.args.get("zone"))
873                )
874
875            return super().attimezone_sql(expression)
def trycast_sql(self, expression: sqlglot.expressions.TryCast) -> str:
877        def trycast_sql(self, expression: exp.TryCast) -> str:
878            return self.cast_sql(expression, safe_prefix="SAFE_")
def bracket_sql(self, expression: sqlglot.expressions.Bracket) -> str:
880        def bracket_sql(self, expression: exp.Bracket) -> str:
881            this = expression.this
882            expressions = expression.expressions
883
884            if len(expressions) == 1 and this and this.is_type(exp.DataType.Type.STRUCT):
885                arg = expressions[0]
886                if arg.type is None:
887                    from sqlglot.optimizer.annotate_types import annotate_types
888
889                    arg = annotate_types(arg)
890
891                if arg.type and arg.type.this in exp.DataType.TEXT_TYPES:
892                    # BQ doesn't support bracket syntax with string values for structs
893                    return f"{self.sql(this)}.{arg.name}"
894
895            expressions_sql = self.expressions(expression, flat=True)
896            offset = expression.args.get("offset")
897
898            if offset == 0:
899                expressions_sql = f"OFFSET({expressions_sql})"
900            elif offset == 1:
901                expressions_sql = f"ORDINAL({expressions_sql})"
902            elif offset is not None:
903                self.unsupported(f"Unsupported array offset: {offset}")
904
905            if expression.args.get("safe"):
906                expressions_sql = f"SAFE_{expressions_sql}"
907
908            return f"{self.sql(this)}[{expressions_sql}]"
def in_unnest_op(self, expression: sqlglot.expressions.Unnest) -> str:
910        def in_unnest_op(self, expression: exp.Unnest) -> str:
911            return self.sql(expression)
def except_op(self, expression: sqlglot.expressions.Except) -> str:
913        def except_op(self, expression: exp.Except) -> str:
914            if not expression.args.get("distinct"):
915                self.unsupported("EXCEPT without DISTINCT is not supported in BigQuery")
916            return f"EXCEPT{' DISTINCT' if expression.args.get('distinct') else ' ALL'}"
def intersect_op(self, expression: sqlglot.expressions.Intersect) -> str:
918        def intersect_op(self, expression: exp.Intersect) -> str:
919            if not expression.args.get("distinct"):
920                self.unsupported("INTERSECT without DISTINCT is not supported in BigQuery")
921            return f"INTERSECT{' DISTINCT' if expression.args.get('distinct') else ' ALL'}"
def version_sql(self, expression: sqlglot.expressions.Version) -> str:
923        def version_sql(self, expression: exp.Version) -> str:
924            if expression.name == "TIMESTAMP":
925                expression.set("this", "SYSTEM_TIME")
926            return super().version_sql(expression)
TRY_SUPPORTED = False
SUPPORTS_UESCAPE = False
AFTER_HAVING_MODIFIER_TRANSFORMS = {'qualify': <function Generator.<lambda>>, 'windows': <function Generator.<lambda>>}
Inherited Members
sqlglot.generator.Generator
Generator
LOCKING_READS_SUPPORTED
WRAP_DERIVED_VALUES
CREATE_FUNCTION_RETURN_AS
MATCHED_BY_SOURCE
SINGLE_STRING_INTERVAL
GROUPINGS_SEP
INDEX_ON
QUERY_HINT_SEP
IS_BOOL_ALLOWED
DUPLICATE_KEY_UPDATE_WITH_SET
LIMIT_IS_TOP
RETURNING_END
EXTRACT_ALLOWS_QUOTES
TZ_TO_WITH_TIME_ZONE
SELECT_KINDS
VALUES_AS_TABLE
ALTER_TABLE_INCLUDE_COLUMN_KEYWORD
AGGREGATE_FILTER_SUPPORTED
SEMI_ANTI_JOIN_WITH_SIDE
COMPUTED_COLUMN_WITH_TYPE
SUPPORTS_TABLE_COPY
TABLESAMPLE_REQUIRES_PARENS
TABLESAMPLE_SIZE_IS_ROWS
TABLESAMPLE_KEYWORDS
TABLESAMPLE_WITH_METHOD
TABLESAMPLE_SEED_KEYWORD
DATA_TYPE_SPECIFIERS_ALLOWED
ENSURE_BOOLS
CTE_RECURSIVE_KEYWORD_REQUIRED
SUPPORTS_SINGLE_ARG_CONCAT
LAST_DAY_SUPPORTS_DATE_PART
INSERT_OVERWRITE
SUPPORTS_SELECT_INTO
SUPPORTS_UNLOGGED_TABLES
SUPPORTS_CREATE_TABLE_LIKE
LIKE_PROPERTY_INSIDE_SCHEMA
MULTI_ARG_DISTINCT
JSON_TYPE_REQUIRED_FOR_EXTRACTION
JSON_PATH_BRACKETED_KEY_SUPPORTED
SET_OP_MODIFIERS
COPY_PARAMS_ARE_WRAPPED
COPY_PARAMS_EQ_REQUIRED
COPY_HAS_INTO_KEYWORD
STAR_EXCEPT
QUOTE_JSON_PATH
TIME_PART_SINGULARS
TOKEN_MAPPING
STRUCT_DELIMITER
PARAMETER_TOKEN
WITH_SEPARATED_COMMENTS
EXCLUDE_COMMENTS
UNWRAPPED_INTERVAL_VALUES
PARAMETERIZABLE_TEXT_TYPES
EXPRESSIONS_WITHOUT_NESTED_CTES
SENTINEL_LINE_BREAK
pretty
identify
normalize
pad
unsupported_level
max_unsupported
leading_comma
max_text_width
comments
dialect
normalize_functions
unsupported_messages
generate
preprocess
unsupported
sep
seg
pad_comment
maybe_comment
wrap
no_identify
normalize_func
indent
sql
uncache_sql
cache_sql
characterset_sql
column_sql
columnposition_sql
columndef_sql
columnconstraint_sql
computedcolumnconstraint_sql
autoincrementcolumnconstraint_sql
compresscolumnconstraint_sql
generatedasidentitycolumnconstraint_sql
generatedasrowcolumnconstraint_sql
periodforsystemtimeconstraint_sql
notnullcolumnconstraint_sql
transformcolumnconstraint_sql
primarykeycolumnconstraint_sql
uniquecolumnconstraint_sql
createable_sql
create_sql
sequenceproperties_sql
clone_sql
describe_sql
heredoc_sql
prepend_ctes
with_sql
cte_sql
tablealias_sql
bitstring_sql
hexstring_sql
bytestring_sql
unicodestring_sql
rawstring_sql
datatypeparam_sql
datatype_sql
directory_sql
delete_sql
drop_sql
except_sql
fetch_sql
filter_sql
hint_sql
indexparameters_sql
index_sql
identifier_sql
hex_sql
lowerhex_sql
inputoutputformat_sql
national_sql
partition_sql
properties_sql
root_properties
properties
with_properties
locate_properties
property_name
property_sql
likeproperty_sql
fallbackproperty_sql
journalproperty_sql
freespaceproperty_sql
checksumproperty_sql
mergeblockratioproperty_sql
datablocksizeproperty_sql
blockcompressionproperty_sql
isolatedloadingproperty_sql
partitionboundspec_sql
partitionedofproperty_sql
lockingproperty_sql
withdataproperty_sql
withsystemversioningproperty_sql
insert_sql
intersect_sql
introducer_sql
kill_sql
pseudotype_sql
objectidentifier_sql
onconflict_sql
returning_sql
rowformatdelimitedproperty_sql
withtablehint_sql
indextablehint_sql
historicaldata_sql
table_sql
tablesample_sql
pivot_sql
tuple_sql
update_sql
values_sql
var_sql
into_sql
from_sql
group_sql
having_sql
connect_sql
prior_sql
join_sql
lambda_sql
lateral_op
lateral_sql
limit_sql
offset_sql
setitem_sql
set_sql
pragma_sql
lock_sql
literal_sql
escape_str
loaddata_sql
null_sql
boolean_sql
order_sql
withfill_sql
cluster_sql
distribute_sql
sort_sql
ordered_sql
matchrecognizemeasure_sql
matchrecognize_sql
query_modifiers
options_modifier
queryoption_sql
offset_limit_modifiers
after_limit_modifiers
select_sql
schema_sql
schema_columns_sql
star_sql
parameter_sql
sessionparameter_sql
placeholder_sql
subquery_sql
qualify_sql
set_operations
union_sql
union_op
unnest_sql
prewhere_sql
where_sql
window_sql
partition_by_sql
windowspec_sql
withingroup_sql
between_sql
bracket_offset_expressions
all_sql
any_sql
exists_sql
case_sql
constraint_sql
nextvaluefor_sql
extract_sql
trim_sql
convert_concat_args
concat_sql
concatws_sql
check_sql
foreignkey_sql
primarykey_sql
if_sql
matchagainst_sql
jsonkeyvalue_sql
jsonpath_sql
json_path_part
formatjson_sql
jsonobject_sql
jsonobjectagg_sql
jsonarray_sql
jsonarrayagg_sql
jsoncolumndef_sql
jsonschema_sql
jsontable_sql
openjsoncolumndef_sql
openjson_sql
in_sql
interval_sql
return_sql
reference_sql
anonymous_sql
paren_sql
neg_sql
not_sql
alias_sql
pivotalias_sql
aliases_sql
atindex_sql
fromtimezone_sql
add_sql
and_sql
or_sql
xor_sql
connector_sql
bitwiseand_sql
bitwiseleftshift_sql
bitwisenot_sql
bitwiseor_sql
bitwiserightshift_sql
bitwisexor_sql
cast_sql
currentdate_sql
currenttimestamp_sql
collate_sql
command_sql
comment_sql
mergetreettlaction_sql
mergetreettl_sql
transaction_sql
commit_sql
rollback_sql
altercolumn_sql
alterdiststyle_sql
altersortkey_sql
renametable_sql
renamecolumn_sql
alterset_sql
altertable_sql
add_column_sql
droppartition_sql
addconstraint_sql
distinct_sql
ignorenulls_sql
respectnulls_sql
havingmax_sql
intdiv_sql
dpipe_sql
div_sql
overlaps_sql
distance_sql
dot_sql
propertyeq_sql
escape_sql
glob_sql
gt_sql
gte_sql
ilike_sql
ilikeany_sql
is_sql
like_sql
likeany_sql
similarto_sql
lt_sql
lte_sql
mul_sql
neq_sql
nullsafeeq_sql
nullsafeneq_sql
slice_sql
sub_sql
try_sql
log_sql
use_sql
binary
function_fallback_sql
func
format_args
too_wide
format_time
expressions
op_expressions
naked_property
tag_sql
token_sql
userdefinedfunction_sql
joinhint_sql
kwarg_sql
when_sql
merge_sql
tochar_sql
tonumber_sql
dictproperty_sql
dictrange_sql
dictsubproperty_sql
oncluster_sql
clusteredbyproperty_sql
anyvalue_sql
querytransform_sql
indexconstraintoption_sql
checkcolumnconstraint_sql
indexcolumnconstraint_sql
nvl2_sql
comprehension_sql
columnprefix_sql
opclass_sql
predict_sql
forin_sql
refresh_sql
operator_sql
toarray_sql
tsordstotime_sql
tsordstotimestamp_sql
tsordstodate_sql
unixdate_sql
lastday_sql
dateadd_sql
arrayany_sql
generateseries_sql
struct_sql
partitionrange_sql
truncatetable_sql
convert_sql
copyparameter_sql
credentials_sql
copy_sql
semicolon_sql
datadeletionproperty_sql
maskingpolicycolumnconstraint_sql
gapfill_sql