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_sql,
 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    timestrtotime_sql,
 26    ts_or_ds_add_cast,
 27)
 28from sqlglot.helper import seq_get, split_num_words
 29from sqlglot.tokens import TokenType
 30
 31if t.TYPE_CHECKING:
 32    from sqlglot._typing import E, Lit
 33
 34logger = logging.getLogger("sqlglot")
 35
 36
 37def _derived_table_values_to_unnest(self: BigQuery.Generator, expression: exp.Values) -> str:
 38    if not expression.find_ancestor(exp.From, exp.Join):
 39        return self.values_sql(expression)
 40
 41    structs = []
 42    alias = expression.args.get("alias")
 43    for tup in expression.find_all(exp.Tuple):
 44        field_aliases = (
 45            alias.columns
 46            if alias and alias.columns
 47            else (f"_c{i}" for i in range(len(tup.expressions)))
 48        )
 49        expressions = [
 50            exp.PropertyEQ(this=exp.to_identifier(name), expression=fld)
 51            for name, fld in zip(field_aliases, tup.expressions)
 52        ]
 53        structs.append(exp.Struct(expressions=expressions))
 54
 55    # Due to `UNNEST_COLUMN_ONLY`, it is expected that the table alias be contained in the columns expression
 56    alias_name_only = exp.TableAlias(columns=[alias.this]) if alias else None
 57    return self.unnest_sql(
 58        exp.Unnest(expressions=[exp.array(*structs, copy=False)], alias=alias_name_only)
 59    )
 60
 61
 62def _returnsproperty_sql(self: BigQuery.Generator, expression: exp.ReturnsProperty) -> str:
 63    this = expression.this
 64    if isinstance(this, exp.Schema):
 65        this = f"{self.sql(this, 'this')} <{self.expressions(this)}>"
 66    else:
 67        this = self.sql(this)
 68    return f"RETURNS {this}"
 69
 70
 71def _create_sql(self: BigQuery.Generator, expression: exp.Create) -> str:
 72    returns = expression.find(exp.ReturnsProperty)
 73    if expression.kind == "FUNCTION" and returns and returns.args.get("is_table"):
 74        expression.set("kind", "TABLE FUNCTION")
 75
 76        if isinstance(expression.expression, (exp.Subquery, exp.Literal)):
 77            expression.set("expression", expression.expression.this)
 78
 79    return self.create_sql(expression)
 80
 81
 82def _unqualify_unnest(expression: exp.Expression) -> exp.Expression:
 83    """Remove references to unnest table aliases since bigquery doesn't allow them.
 84
 85    These are added by the optimizer's qualify_column step.
 86    """
 87    from sqlglot.optimizer.scope import find_all_in_scope
 88
 89    if isinstance(expression, exp.Select):
 90        unnest_aliases = {
 91            unnest.alias
 92            for unnest in find_all_in_scope(expression, exp.Unnest)
 93            if isinstance(unnest.parent, (exp.From, exp.Join))
 94        }
 95        if unnest_aliases:
 96            for column in expression.find_all(exp.Column):
 97                if column.table in unnest_aliases:
 98                    column.set("table", None)
 99                elif column.db in unnest_aliases:
100                    column.set("db", None)
101
102    return expression
103
104
105# https://issuetracker.google.com/issues/162294746
106# workaround for bigquery bug when grouping by an expression and then ordering
107# WITH x AS (SELECT 1 y)
108# SELECT y + 1 z
109# FROM x
110# GROUP BY x + 1
111# ORDER by z
112def _alias_ordered_group(expression: exp.Expression) -> exp.Expression:
113    if isinstance(expression, exp.Select):
114        group = expression.args.get("group")
115        order = expression.args.get("order")
116
117        if group and order:
118            aliases = {
119                select.this: select.args["alias"]
120                for select in expression.selects
121                if isinstance(select, exp.Alias)
122            }
123
124            for grouped in group.expressions:
125                if grouped.is_int:
126                    continue
127                alias = aliases.get(grouped)
128                if alias:
129                    grouped.replace(exp.column(alias))
130
131    return expression
132
133
134def _pushdown_cte_column_names(expression: exp.Expression) -> exp.Expression:
135    """BigQuery doesn't allow column names when defining a CTE, so we try to push them down."""
136    if isinstance(expression, exp.CTE) and expression.alias_column_names:
137        cte_query = expression.this
138
139        if cte_query.is_star:
140            logger.warning(
141                "Can't push down CTE column names for star queries. Run the query through"
142                " the optimizer or use 'qualify' to expand the star projections first."
143            )
144            return expression
145
146        column_names = expression.alias_column_names
147        expression.args["alias"].set("columns", None)
148
149        for name, select in zip(column_names, cte_query.selects):
150            to_replace = select
151
152            if isinstance(select, exp.Alias):
153                select = select.this
154
155            # Inner aliases are shadowed by the CTE column names
156            to_replace.replace(exp.alias_(select, name))
157
158    return expression
159
160
161def _build_parse_timestamp(args: t.List) -> exp.StrToTime:
162    this = build_formatted_time(exp.StrToTime, "bigquery")([seq_get(args, 1), seq_get(args, 0)])
163    this.set("zone", seq_get(args, 2))
164    return this
165
166
167def _build_timestamp(args: t.List) -> exp.Timestamp:
168    timestamp = exp.Timestamp.from_arg_list(args)
169    timestamp.set("with_tz", True)
170    return timestamp
171
172
173def _build_date(args: t.List) -> exp.Date | exp.DateFromParts:
174    expr_type = exp.DateFromParts if len(args) == 3 else exp.Date
175    return expr_type.from_arg_list(args)
176
177
178def _build_to_hex(args: t.List) -> exp.Hex | exp.MD5:
179    # TO_HEX(MD5(..)) is common in BigQuery, so it's parsed into MD5 to simplify its transpilation
180    arg = seq_get(args, 0)
181    return exp.MD5(this=arg.this) if isinstance(arg, exp.MD5Digest) else exp.Hex(this=arg)
182
183
184def _array_contains_sql(self: BigQuery.Generator, expression: exp.ArrayContains) -> str:
185    return self.sql(
186        exp.Exists(
187            this=exp.select("1")
188            .from_(exp.Unnest(expressions=[expression.left]).as_("_unnest", table=["_col"]))
189            .where(exp.column("_col").eq(expression.right))
190        )
191    )
192
193
194def _ts_or_ds_add_sql(self: BigQuery.Generator, expression: exp.TsOrDsAdd) -> str:
195    return date_add_interval_sql("DATE", "ADD")(self, ts_or_ds_add_cast(expression))
196
197
198def _ts_or_ds_diff_sql(self: BigQuery.Generator, expression: exp.TsOrDsDiff) -> str:
199    expression.this.replace(exp.cast(expression.this, "TIMESTAMP", copy=True))
200    expression.expression.replace(exp.cast(expression.expression, "TIMESTAMP", copy=True))
201    unit = expression.args.get("unit") or "DAY"
202    return self.func("DATE_DIFF", expression.this, expression.expression, unit)
203
204
205def _unix_to_time_sql(self: BigQuery.Generator, expression: exp.UnixToTime) -> str:
206    scale = expression.args.get("scale")
207    timestamp = expression.this
208
209    if scale in (None, exp.UnixToTime.SECONDS):
210        return self.func("TIMESTAMP_SECONDS", timestamp)
211    if scale == exp.UnixToTime.MILLIS:
212        return self.func("TIMESTAMP_MILLIS", timestamp)
213    if scale == exp.UnixToTime.MICROS:
214        return self.func("TIMESTAMP_MICROS", timestamp)
215
216    unix_seconds = exp.cast(exp.Div(this=timestamp, expression=exp.func("POW", 10, scale)), "int64")
217    return self.func("TIMESTAMP_SECONDS", unix_seconds)
218
219
220def _build_time(args: t.List) -> exp.Func:
221    if len(args) == 1:
222        return exp.TsOrDsToTime(this=args[0])
223    if len(args) == 3:
224        return exp.TimeFromParts.from_arg_list(args)
225
226    return exp.Anonymous(this="TIME", expressions=args)
227
228
229class BigQuery(Dialect):
230    WEEK_OFFSET = -1
231    UNNEST_COLUMN_ONLY = True
232    SUPPORTS_USER_DEFINED_TYPES = False
233    SUPPORTS_SEMI_ANTI_JOIN = False
234    LOG_BASE_FIRST = False
235
236    # https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#case_sensitivity
237    NORMALIZATION_STRATEGY = NormalizationStrategy.CASE_INSENSITIVE
238
239    # bigquery udfs are case sensitive
240    NORMALIZE_FUNCTIONS = False
241
242    # https://cloud.google.com/bigquery/docs/reference/standard-sql/format-elements#format_elements_date_time
243    TIME_MAPPING = {
244        "%D": "%m/%d/%y",
245        "%E*S": "%S.%f",
246        "%E6S": "%S.%f",
247    }
248
249    FORMAT_MAPPING = {
250        "DD": "%d",
251        "MM": "%m",
252        "MON": "%b",
253        "MONTH": "%B",
254        "YYYY": "%Y",
255        "YY": "%y",
256        "HH": "%I",
257        "HH12": "%I",
258        "HH24": "%H",
259        "MI": "%M",
260        "SS": "%S",
261        "SSSSS": "%f",
262        "TZH": "%z",
263    }
264
265    # The _PARTITIONTIME and _PARTITIONDATE pseudo-columns are not returned by a SELECT * statement
266    # https://cloud.google.com/bigquery/docs/querying-partitioned-tables#query_an_ingestion-time_partitioned_table
267    PSEUDOCOLUMNS = {"_PARTITIONTIME", "_PARTITIONDATE"}
268
269    def normalize_identifier(self, expression: E) -> E:
270        if isinstance(expression, exp.Identifier):
271            parent = expression.parent
272            while isinstance(parent, exp.Dot):
273                parent = parent.parent
274
275            # In BigQuery, CTEs are case-insensitive, but UDF and table names are case-sensitive
276            # by default. The following check uses a heuristic to detect tables based on whether
277            # they are qualified. This should generally be correct, because tables in BigQuery
278            # must be qualified with at least a dataset, unless @@dataset_id is set.
279            case_sensitive = (
280                isinstance(parent, exp.UserDefinedFunction)
281                or (
282                    isinstance(parent, exp.Table)
283                    and parent.db
284                    and (parent.meta.get("quoted_table") or not parent.meta.get("maybe_column"))
285                )
286                or expression.meta.get("is_table")
287            )
288            if not case_sensitive:
289                expression.set("this", expression.this.lower())
290
291        return expression
292
293    class Tokenizer(tokens.Tokenizer):
294        QUOTES = ["'", '"', '"""', "'''"]
295        COMMENTS = ["--", "#", ("/*", "*/")]
296        IDENTIFIERS = ["`"]
297        STRING_ESCAPES = ["\\"]
298
299        HEX_STRINGS = [("0x", ""), ("0X", "")]
300
301        BYTE_STRINGS = [
302            (prefix + q, q) for q in t.cast(t.List[str], QUOTES) for prefix in ("b", "B")
303        ]
304
305        RAW_STRINGS = [
306            (prefix + q, q) for q in t.cast(t.List[str], QUOTES) for prefix in ("r", "R")
307        ]
308
309        KEYWORDS = {
310            **tokens.Tokenizer.KEYWORDS,
311            "ANY TYPE": TokenType.VARIANT,
312            "BEGIN": TokenType.COMMAND,
313            "BEGIN TRANSACTION": TokenType.BEGIN,
314            "BYTES": TokenType.BINARY,
315            "CURRENT_DATETIME": TokenType.CURRENT_DATETIME,
316            "DATETIME": TokenType.TIMESTAMP,
317            "DECLARE": TokenType.COMMAND,
318            "ELSEIF": TokenType.COMMAND,
319            "EXCEPTION": TokenType.COMMAND,
320            "FLOAT64": TokenType.DOUBLE,
321            "FOR SYSTEM_TIME": TokenType.TIMESTAMP_SNAPSHOT,
322            "MODEL": TokenType.MODEL,
323            "NOT DETERMINISTIC": TokenType.VOLATILE,
324            "RECORD": TokenType.STRUCT,
325            "TIMESTAMP": TokenType.TIMESTAMPTZ,
326        }
327        KEYWORDS.pop("DIV")
328        KEYWORDS.pop("VALUES")
329
330    class Parser(parser.Parser):
331        PREFIXED_PIVOT_COLUMNS = True
332        LOG_DEFAULTS_TO_LN = True
333        SUPPORTS_IMPLICIT_UNNEST = True
334
335        FUNCTIONS = {
336            **parser.Parser.FUNCTIONS,
337            "DATE": _build_date,
338            "DATE_ADD": build_date_delta_with_interval(exp.DateAdd),
339            "DATE_SUB": build_date_delta_with_interval(exp.DateSub),
340            "DATE_TRUNC": lambda args: exp.DateTrunc(
341                unit=exp.Literal.string(str(seq_get(args, 1))),
342                this=seq_get(args, 0),
343            ),
344            "DATETIME_ADD": build_date_delta_with_interval(exp.DatetimeAdd),
345            "DATETIME_SUB": build_date_delta_with_interval(exp.DatetimeSub),
346            "DIV": binary_from_function(exp.IntDiv),
347            "FORMAT_DATE": lambda args: exp.TimeToStr(
348                this=exp.TsOrDsToDate(this=seq_get(args, 1)), format=seq_get(args, 0)
349            ),
350            "GENERATE_ARRAY": exp.GenerateSeries.from_arg_list,
351            "JSON_EXTRACT_SCALAR": lambda args: exp.JSONExtractScalar(
352                this=seq_get(args, 0), expression=seq_get(args, 1) or exp.Literal.string("$")
353            ),
354            "MD5": exp.MD5Digest.from_arg_list,
355            "TO_HEX": _build_to_hex,
356            "PARSE_DATE": lambda args: build_formatted_time(exp.StrToDate, "bigquery")(
357                [seq_get(args, 1), seq_get(args, 0)]
358            ),
359            "PARSE_TIMESTAMP": _build_parse_timestamp,
360            "REGEXP_CONTAINS": exp.RegexpLike.from_arg_list,
361            "REGEXP_EXTRACT": lambda args: exp.RegexpExtract(
362                this=seq_get(args, 0),
363                expression=seq_get(args, 1),
364                position=seq_get(args, 2),
365                occurrence=seq_get(args, 3),
366                group=exp.Literal.number(1) if re.compile(args[1].name).groups == 1 else None,
367            ),
368            "SHA256": lambda args: exp.SHA2(this=seq_get(args, 0), length=exp.Literal.number(256)),
369            "SHA512": lambda args: exp.SHA2(this=seq_get(args, 0), length=exp.Literal.number(512)),
370            "SPLIT": lambda args: exp.Split(
371                # https://cloud.google.com/bigquery/docs/reference/standard-sql/string_functions#split
372                this=seq_get(args, 0),
373                expression=seq_get(args, 1) or exp.Literal.string(","),
374            ),
375            "TIME": _build_time,
376            "TIME_ADD": build_date_delta_with_interval(exp.TimeAdd),
377            "TIME_SUB": build_date_delta_with_interval(exp.TimeSub),
378            "TIMESTAMP": _build_timestamp,
379            "TIMESTAMP_ADD": build_date_delta_with_interval(exp.TimestampAdd),
380            "TIMESTAMP_SUB": build_date_delta_with_interval(exp.TimestampSub),
381            "TIMESTAMP_MICROS": lambda args: exp.UnixToTime(
382                this=seq_get(args, 0), scale=exp.UnixToTime.MICROS
383            ),
384            "TIMESTAMP_MILLIS": lambda args: exp.UnixToTime(
385                this=seq_get(args, 0), scale=exp.UnixToTime.MILLIS
386            ),
387            "TIMESTAMP_SECONDS": lambda args: exp.UnixToTime(this=seq_get(args, 0)),
388            "TO_JSON_STRING": exp.JSONFormat.from_arg_list,
389        }
390
391        FUNCTION_PARSERS = {
392            **parser.Parser.FUNCTION_PARSERS,
393            "ARRAY": lambda self: self.expression(exp.Array, expressions=[self._parse_statement()]),
394        }
395        FUNCTION_PARSERS.pop("TRIM")
396
397        NO_PAREN_FUNCTIONS = {
398            **parser.Parser.NO_PAREN_FUNCTIONS,
399            TokenType.CURRENT_DATETIME: exp.CurrentDatetime,
400        }
401
402        NESTED_TYPE_TOKENS = {
403            *parser.Parser.NESTED_TYPE_TOKENS,
404            TokenType.TABLE,
405        }
406
407        PROPERTY_PARSERS = {
408            **parser.Parser.PROPERTY_PARSERS,
409            "NOT DETERMINISTIC": lambda self: self.expression(
410                exp.StabilityProperty, this=exp.Literal.string("VOLATILE")
411            ),
412            "OPTIONS": lambda self: self._parse_with_property(),
413        }
414
415        CONSTRAINT_PARSERS = {
416            **parser.Parser.CONSTRAINT_PARSERS,
417            "OPTIONS": lambda self: exp.Properties(expressions=self._parse_with_property()),
418        }
419
420        RANGE_PARSERS = parser.Parser.RANGE_PARSERS.copy()
421        RANGE_PARSERS.pop(TokenType.OVERLAPS)
422
423        NULL_TOKENS = {TokenType.NULL, TokenType.UNKNOWN}
424
425        STATEMENT_PARSERS = {
426            **parser.Parser.STATEMENT_PARSERS,
427            TokenType.ELSE: lambda self: self._parse_as_command(self._prev),
428            TokenType.END: lambda self: self._parse_as_command(self._prev),
429            TokenType.FOR: lambda self: self._parse_for_in(),
430        }
431
432        BRACKET_OFFSETS = {
433            "OFFSET": (0, False),
434            "ORDINAL": (1, False),
435            "SAFE_OFFSET": (0, True),
436            "SAFE_ORDINAL": (1, True),
437        }
438
439        def _parse_for_in(self) -> exp.ForIn:
440            this = self._parse_range()
441            self._match_text_seq("DO")
442            return self.expression(exp.ForIn, this=this, expression=self._parse_statement())
443
444        def _parse_table_part(self, schema: bool = False) -> t.Optional[exp.Expression]:
445            this = super()._parse_table_part(schema=schema) or self._parse_number()
446
447            # https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#table_names
448            if isinstance(this, exp.Identifier):
449                table_name = this.name
450                while self._match(TokenType.DASH, advance=False) and self._next:
451                    text = ""
452                    while self._curr and self._curr.token_type != TokenType.DOT:
453                        self._advance()
454                        text += self._prev.text
455                    table_name += text
456
457                this = exp.Identifier(this=table_name, quoted=this.args.get("quoted"))
458            elif isinstance(this, exp.Literal):
459                table_name = this.name
460
461                if self._is_connected() and self._parse_var(any_token=True):
462                    table_name += self._prev.text
463
464                this = exp.Identifier(this=table_name, quoted=True)
465
466            return this
467
468        def _parse_table_parts(
469            self, schema: bool = False, is_db_reference: bool = False, wildcard: bool = False
470        ) -> exp.Table:
471            table = super()._parse_table_parts(
472                schema=schema, is_db_reference=is_db_reference, wildcard=True
473            )
474
475            # proj-1.db.tbl -- `1.` is tokenized as a float so we need to unravel it here
476            if not table.catalog:
477                if table.db:
478                    parts = table.db.split(".")
479                    if len(parts) == 2 and not table.args["db"].quoted:
480                        table.set("catalog", exp.Identifier(this=parts[0]))
481                        table.set("db", exp.Identifier(this=parts[1]))
482                else:
483                    parts = table.name.split(".")
484                    if len(parts) == 2 and not table.this.quoted:
485                        table.set("db", exp.Identifier(this=parts[0]))
486                        table.set("this", exp.Identifier(this=parts[1]))
487
488            if any("." in p.name for p in table.parts):
489                catalog, db, this, *rest = (
490                    exp.to_identifier(p, quoted=True)
491                    for p in split_num_words(".".join(p.name for p in table.parts), ".", 3)
492                )
493
494                if rest and this:
495                    this = exp.Dot.build([this, *rest])  # type: ignore
496
497                table = exp.Table(this=this, db=db, catalog=catalog)
498                table.meta["quoted_table"] = True
499
500            return table
501
502        @t.overload
503        def _parse_json_object(self, agg: Lit[False]) -> exp.JSONObject: ...
504
505        @t.overload
506        def _parse_json_object(self, agg: Lit[True]) -> exp.JSONObjectAgg: ...
507
508        def _parse_json_object(self, agg=False):
509            json_object = super()._parse_json_object()
510            array_kv_pair = seq_get(json_object.expressions, 0)
511
512            # Converts BQ's "signature 2" of JSON_OBJECT into SQLGlot's canonical representation
513            # https://cloud.google.com/bigquery/docs/reference/standard-sql/json_functions#json_object_signature2
514            if (
515                array_kv_pair
516                and isinstance(array_kv_pair.this, exp.Array)
517                and isinstance(array_kv_pair.expression, exp.Array)
518            ):
519                keys = array_kv_pair.this.expressions
520                values = array_kv_pair.expression.expressions
521
522                json_object.set(
523                    "expressions",
524                    [exp.JSONKeyValue(this=k, expression=v) for k, v in zip(keys, values)],
525                )
526
527            return json_object
528
529        def _parse_bracket(
530            self, this: t.Optional[exp.Expression] = None
531        ) -> t.Optional[exp.Expression]:
532            bracket = super()._parse_bracket(this)
533
534            if this is bracket:
535                return bracket
536
537            if isinstance(bracket, exp.Bracket):
538                for expression in bracket.expressions:
539                    name = expression.name.upper()
540
541                    if name not in self.BRACKET_OFFSETS:
542                        break
543
544                    offset, safe = self.BRACKET_OFFSETS[name]
545                    bracket.set("offset", offset)
546                    bracket.set("safe", safe)
547                    expression.replace(expression.expressions[0])
548
549            return bracket
550
551    class Generator(generator.Generator):
552        EXPLICIT_UNION = True
553        INTERVAL_ALLOWS_PLURAL_FORM = False
554        JOIN_HINTS = False
555        QUERY_HINTS = False
556        TABLE_HINTS = False
557        LIMIT_FETCH = "LIMIT"
558        RENAME_TABLE_WITH_DB = False
559        NVL2_SUPPORTED = False
560        UNNEST_WITH_ORDINALITY = False
561        COLLATE_IS_FUNC = True
562        LIMIT_ONLY_LITERALS = True
563        SUPPORTS_TABLE_ALIAS_COLUMNS = False
564        UNPIVOT_ALIASES_ARE_IDENTIFIERS = False
565        JSON_KEY_VALUE_PAIR_SEP = ","
566        NULL_ORDERING_SUPPORTED = False
567        IGNORE_NULLS_IN_FUNC = True
568        JSON_PATH_SINGLE_QUOTE_ESCAPE = True
569        CAN_IMPLEMENT_ARRAY_ANY = True
570        SUPPORTS_TO_NUMBER = False
571        NAMED_PLACEHOLDER_TOKEN = "@"
572
573        TRANSFORMS = {
574            **generator.Generator.TRANSFORMS,
575            exp.ApproxDistinct: rename_func("APPROX_COUNT_DISTINCT"),
576            exp.ArgMax: arg_max_or_min_no_count("MAX_BY"),
577            exp.ArgMin: arg_max_or_min_no_count("MIN_BY"),
578            exp.ArrayContains: _array_contains_sql,
579            exp.ArrayFilter: filter_array_using_unnest,
580            exp.ArraySize: rename_func("ARRAY_LENGTH"),
581            exp.Cast: transforms.preprocess([transforms.remove_precision_parameterized_types]),
582            exp.CollateProperty: lambda self, e: (
583                f"DEFAULT COLLATE {self.sql(e, 'this')}"
584                if e.args.get("default")
585                else f"COLLATE {self.sql(e, 'this')}"
586            ),
587            exp.Commit: lambda *_: "COMMIT TRANSACTION",
588            exp.CountIf: rename_func("COUNTIF"),
589            exp.Create: _create_sql,
590            exp.CTE: transforms.preprocess([_pushdown_cte_column_names]),
591            exp.DateAdd: date_add_interval_sql("DATE", "ADD"),
592            exp.DateDiff: lambda self, e: self.func(
593                "DATE_DIFF", e.this, e.expression, e.unit or "DAY"
594            ),
595            exp.DateFromParts: rename_func("DATE"),
596            exp.DateStrToDate: datestrtodate_sql,
597            exp.DateSub: date_add_interval_sql("DATE", "SUB"),
598            exp.DatetimeAdd: date_add_interval_sql("DATETIME", "ADD"),
599            exp.DatetimeSub: date_add_interval_sql("DATETIME", "SUB"),
600            exp.DateTrunc: lambda self, e: self.func("DATE_TRUNC", e.this, e.text("unit")),
601            exp.FromTimeZone: lambda self, e: self.func(
602                "DATETIME", self.func("TIMESTAMP", e.this, e.args.get("zone")), "'UTC'"
603            ),
604            exp.GenerateSeries: rename_func("GENERATE_ARRAY"),
605            exp.GroupConcat: rename_func("STRING_AGG"),
606            exp.Hex: rename_func("TO_HEX"),
607            exp.If: if_sql(false_value="NULL"),
608            exp.ILike: no_ilike_sql,
609            exp.IntDiv: rename_func("DIV"),
610            exp.JSONFormat: rename_func("TO_JSON_STRING"),
611            exp.Max: max_or_greatest,
612            exp.Mod: rename_func("MOD"),
613            exp.MD5: lambda self, e: self.func("TO_HEX", self.func("MD5", e.this)),
614            exp.MD5Digest: rename_func("MD5"),
615            exp.Min: min_or_least,
616            exp.PartitionedByProperty: lambda self, e: f"PARTITION BY {self.sql(e, 'this')}",
617            exp.RegexpExtract: lambda self, e: self.func(
618                "REGEXP_EXTRACT",
619                e.this,
620                e.expression,
621                e.args.get("position"),
622                e.args.get("occurrence"),
623            ),
624            exp.RegexpReplace: regexp_replace_sql,
625            exp.RegexpLike: rename_func("REGEXP_CONTAINS"),
626            exp.ReturnsProperty: _returnsproperty_sql,
627            exp.Rollback: lambda *_: "ROLLBACK TRANSACTION",
628            exp.Select: transforms.preprocess(
629                [
630                    transforms.explode_to_unnest(),
631                    _unqualify_unnest,
632                    transforms.eliminate_distinct_on,
633                    _alias_ordered_group,
634                    transforms.eliminate_semi_and_anti_joins,
635                ]
636            ),
637            exp.SHA2: lambda self, e: self.func(
638                "SHA256" if e.text("length") == "256" else "SHA512", e.this
639            ),
640            exp.StabilityProperty: lambda self, e: (
641                "DETERMINISTIC" if e.name == "IMMUTABLE" else "NOT DETERMINISTIC"
642            ),
643            exp.StrToDate: lambda self, e: self.func("PARSE_DATE", self.format_time(e), e.this),
644            exp.StrToTime: lambda self, e: self.func(
645                "PARSE_TIMESTAMP", self.format_time(e), e.this, e.args.get("zone")
646            ),
647            exp.TimeAdd: date_add_interval_sql("TIME", "ADD"),
648            exp.TimeFromParts: rename_func("TIME"),
649            exp.TimeSub: date_add_interval_sql("TIME", "SUB"),
650            exp.TimestampAdd: date_add_interval_sql("TIMESTAMP", "ADD"),
651            exp.TimestampDiff: rename_func("TIMESTAMP_DIFF"),
652            exp.TimestampSub: date_add_interval_sql("TIMESTAMP", "SUB"),
653            exp.TimeStrToTime: timestrtotime_sql,
654            exp.Transaction: lambda *_: "BEGIN TRANSACTION",
655            exp.Trim: lambda self, e: self.func("TRIM", e.this, e.expression),
656            exp.TsOrDsAdd: _ts_or_ds_add_sql,
657            exp.TsOrDsDiff: _ts_or_ds_diff_sql,
658            exp.TsOrDsToTime: rename_func("TIME"),
659            exp.Unhex: rename_func("FROM_HEX"),
660            exp.UnixDate: rename_func("UNIX_DATE"),
661            exp.UnixToTime: _unix_to_time_sql,
662            exp.Values: _derived_table_values_to_unnest,
663            exp.VariancePop: rename_func("VAR_POP"),
664        }
665
666        SUPPORTED_JSON_PATH_PARTS = {
667            exp.JSONPathKey,
668            exp.JSONPathRoot,
669            exp.JSONPathSubscript,
670        }
671
672        TYPE_MAPPING = {
673            **generator.Generator.TYPE_MAPPING,
674            exp.DataType.Type.BIGDECIMAL: "BIGNUMERIC",
675            exp.DataType.Type.BIGINT: "INT64",
676            exp.DataType.Type.BINARY: "BYTES",
677            exp.DataType.Type.BOOLEAN: "BOOL",
678            exp.DataType.Type.CHAR: "STRING",
679            exp.DataType.Type.DECIMAL: "NUMERIC",
680            exp.DataType.Type.DOUBLE: "FLOAT64",
681            exp.DataType.Type.FLOAT: "FLOAT64",
682            exp.DataType.Type.INT: "INT64",
683            exp.DataType.Type.NCHAR: "STRING",
684            exp.DataType.Type.NVARCHAR: "STRING",
685            exp.DataType.Type.SMALLINT: "INT64",
686            exp.DataType.Type.TEXT: "STRING",
687            exp.DataType.Type.TIMESTAMP: "DATETIME",
688            exp.DataType.Type.TIMESTAMPTZ: "TIMESTAMP",
689            exp.DataType.Type.TIMESTAMPLTZ: "TIMESTAMP",
690            exp.DataType.Type.TINYINT: "INT64",
691            exp.DataType.Type.VARBINARY: "BYTES",
692            exp.DataType.Type.VARCHAR: "STRING",
693            exp.DataType.Type.VARIANT: "ANY TYPE",
694        }
695
696        PROPERTIES_LOCATION = {
697            **generator.Generator.PROPERTIES_LOCATION,
698            exp.PartitionedByProperty: exp.Properties.Location.POST_SCHEMA,
699            exp.VolatileProperty: exp.Properties.Location.UNSUPPORTED,
700        }
701
702        # from: https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#reserved_keywords
703        RESERVED_KEYWORDS = {
704            *generator.Generator.RESERVED_KEYWORDS,
705            "all",
706            "and",
707            "any",
708            "array",
709            "as",
710            "asc",
711            "assert_rows_modified",
712            "at",
713            "between",
714            "by",
715            "case",
716            "cast",
717            "collate",
718            "contains",
719            "create",
720            "cross",
721            "cube",
722            "current",
723            "default",
724            "define",
725            "desc",
726            "distinct",
727            "else",
728            "end",
729            "enum",
730            "escape",
731            "except",
732            "exclude",
733            "exists",
734            "extract",
735            "false",
736            "fetch",
737            "following",
738            "for",
739            "from",
740            "full",
741            "group",
742            "grouping",
743            "groups",
744            "hash",
745            "having",
746            "if",
747            "ignore",
748            "in",
749            "inner",
750            "intersect",
751            "interval",
752            "into",
753            "is",
754            "join",
755            "lateral",
756            "left",
757            "like",
758            "limit",
759            "lookup",
760            "merge",
761            "natural",
762            "new",
763            "no",
764            "not",
765            "null",
766            "nulls",
767            "of",
768            "on",
769            "or",
770            "order",
771            "outer",
772            "over",
773            "partition",
774            "preceding",
775            "proto",
776            "qualify",
777            "range",
778            "recursive",
779            "respect",
780            "right",
781            "rollup",
782            "rows",
783            "select",
784            "set",
785            "some",
786            "struct",
787            "tablesample",
788            "then",
789            "to",
790            "treat",
791            "true",
792            "unbounded",
793            "union",
794            "unnest",
795            "using",
796            "when",
797            "where",
798            "window",
799            "with",
800            "within",
801        }
802
803        def table_parts(self, expression: exp.Table) -> str:
804            # Depending on the context, `x.y` may not resolve to the same data source as `x`.`y`, so
805            # we need to make sure the correct quoting is used in each case.
806            #
807            # For example, if there is a CTE x that clashes with a schema name, then the former will
808            # return the table y in that schema, whereas the latter will return the CTE's y column:
809            #
810            # - WITH x AS (SELECT [1, 2] AS y) SELECT * FROM x, `x.y`   -> cross join
811            # - WITH x AS (SELECT [1, 2] AS y) SELECT * FROM x, `x`.`y` -> implicit unnest
812            if expression.meta.get("quoted_table"):
813                table_parts = ".".join(p.name for p in expression.parts)
814                return self.sql(exp.Identifier(this=table_parts, quoted=True))
815
816            return super().table_parts(expression)
817
818        def timetostr_sql(self, expression: exp.TimeToStr) -> str:
819            this = expression.this if isinstance(expression.this, exp.TsOrDsToDate) else expression
820            return self.func("FORMAT_DATE", self.format_time(expression), this.this)
821
822        def eq_sql(self, expression: exp.EQ) -> str:
823            # Operands of = cannot be NULL in BigQuery
824            if isinstance(expression.left, exp.Null) or isinstance(expression.right, exp.Null):
825                if not isinstance(expression.parent, exp.Update):
826                    return "NULL"
827
828            return self.binary(expression, "=")
829
830        def attimezone_sql(self, expression: exp.AtTimeZone) -> str:
831            parent = expression.parent
832
833            # BigQuery allows CAST(.. AS {STRING|TIMESTAMP} [FORMAT <fmt> [AT TIME ZONE <tz>]]).
834            # Only the TIMESTAMP one should use the below conversion, when AT TIME ZONE is included.
835            if not isinstance(parent, exp.Cast) or not parent.to.is_type("text"):
836                return self.func(
837                    "TIMESTAMP", self.func("DATETIME", expression.this, expression.args.get("zone"))
838                )
839
840            return super().attimezone_sql(expression)
841
842        def trycast_sql(self, expression: exp.TryCast) -> str:
843            return self.cast_sql(expression, safe_prefix="SAFE_")
844
845        def array_sql(self, expression: exp.Array) -> str:
846            first_arg = seq_get(expression.expressions, 0)
847            if isinstance(first_arg, exp.Query):
848                return f"ARRAY{self.wrap(self.sql(first_arg))}"
849
850            return inline_array_sql(self, expression)
851
852        def bracket_sql(self, expression: exp.Bracket) -> str:
853            this = expression.this
854            expressions = expression.expressions
855
856            if len(expressions) == 1 and this and this.is_type(exp.DataType.Type.STRUCT):
857                arg = expressions[0]
858                if arg.type is None:
859                    from sqlglot.optimizer.annotate_types import annotate_types
860
861                    arg = annotate_types(arg)
862
863                if arg.type and arg.type.this in exp.DataType.TEXT_TYPES:
864                    # BQ doesn't support bracket syntax with string values for structs
865                    return f"{self.sql(this)}.{arg.name}"
866
867            expressions_sql = self.expressions(expression, flat=True)
868            offset = expression.args.get("offset")
869
870            if offset == 0:
871                expressions_sql = f"OFFSET({expressions_sql})"
872            elif offset == 1:
873                expressions_sql = f"ORDINAL({expressions_sql})"
874            elif offset is not None:
875                self.unsupported(f"Unsupported array offset: {offset}")
876
877            if expression.args.get("safe"):
878                expressions_sql = f"SAFE_{expressions_sql}"
879
880            return f"{self.sql(this)}[{expressions_sql}]"
881
882        def in_unnest_op(self, expression: exp.Unnest) -> str:
883            return self.sql(expression)
884
885        def except_op(self, expression: exp.Except) -> str:
886            if not expression.args.get("distinct"):
887                self.unsupported("EXCEPT without DISTINCT is not supported in BigQuery")
888            return f"EXCEPT{' DISTINCT' if expression.args.get('distinct') else ' ALL'}"
889
890        def intersect_op(self, expression: exp.Intersect) -> str:
891            if not expression.args.get("distinct"):
892                self.unsupported("INTERSECT without DISTINCT is not supported in BigQuery")
893            return f"INTERSECT{' DISTINCT' if expression.args.get('distinct') else ' ALL'}"
894
895        def with_properties(self, properties: exp.Properties) -> str:
896            return self.properties(properties, prefix=self.seg("OPTIONS"))
897
898        def version_sql(self, expression: exp.Version) -> str:
899            if expression.name == "TIMESTAMP":
900                expression.set("this", "SYSTEM_TIME")
901            return super().version_sql(expression)
logger = <Logger sqlglot (WARNING)>
class BigQuery(sqlglot.dialects.dialect.Dialect):
230class BigQuery(Dialect):
231    WEEK_OFFSET = -1
232    UNNEST_COLUMN_ONLY = True
233    SUPPORTS_USER_DEFINED_TYPES = False
234    SUPPORTS_SEMI_ANTI_JOIN = False
235    LOG_BASE_FIRST = False
236
237    # https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#case_sensitivity
238    NORMALIZATION_STRATEGY = NormalizationStrategy.CASE_INSENSITIVE
239
240    # bigquery udfs are case sensitive
241    NORMALIZE_FUNCTIONS = False
242
243    # https://cloud.google.com/bigquery/docs/reference/standard-sql/format-elements#format_elements_date_time
244    TIME_MAPPING = {
245        "%D": "%m/%d/%y",
246        "%E*S": "%S.%f",
247        "%E6S": "%S.%f",
248    }
249
250    FORMAT_MAPPING = {
251        "DD": "%d",
252        "MM": "%m",
253        "MON": "%b",
254        "MONTH": "%B",
255        "YYYY": "%Y",
256        "YY": "%y",
257        "HH": "%I",
258        "HH12": "%I",
259        "HH24": "%H",
260        "MI": "%M",
261        "SS": "%S",
262        "SSSSS": "%f",
263        "TZH": "%z",
264    }
265
266    # The _PARTITIONTIME and _PARTITIONDATE pseudo-columns are not returned by a SELECT * statement
267    # https://cloud.google.com/bigquery/docs/querying-partitioned-tables#query_an_ingestion-time_partitioned_table
268    PSEUDOCOLUMNS = {"_PARTITIONTIME", "_PARTITIONDATE"}
269
270    def normalize_identifier(self, expression: E) -> E:
271        if isinstance(expression, exp.Identifier):
272            parent = expression.parent
273            while isinstance(parent, exp.Dot):
274                parent = parent.parent
275
276            # In BigQuery, CTEs are case-insensitive, but UDF and table names are case-sensitive
277            # by default. The following check uses a heuristic to detect tables based on whether
278            # they are qualified. This should generally be correct, because tables in BigQuery
279            # must be qualified with at least a dataset, unless @@dataset_id is set.
280            case_sensitive = (
281                isinstance(parent, exp.UserDefinedFunction)
282                or (
283                    isinstance(parent, exp.Table)
284                    and parent.db
285                    and (parent.meta.get("quoted_table") or not parent.meta.get("maybe_column"))
286                )
287                or expression.meta.get("is_table")
288            )
289            if not case_sensitive:
290                expression.set("this", expression.this.lower())
291
292        return expression
293
294    class Tokenizer(tokens.Tokenizer):
295        QUOTES = ["'", '"', '"""', "'''"]
296        COMMENTS = ["--", "#", ("/*", "*/")]
297        IDENTIFIERS = ["`"]
298        STRING_ESCAPES = ["\\"]
299
300        HEX_STRINGS = [("0x", ""), ("0X", "")]
301
302        BYTE_STRINGS = [
303            (prefix + q, q) for q in t.cast(t.List[str], QUOTES) for prefix in ("b", "B")
304        ]
305
306        RAW_STRINGS = [
307            (prefix + q, q) for q in t.cast(t.List[str], QUOTES) for prefix in ("r", "R")
308        ]
309
310        KEYWORDS = {
311            **tokens.Tokenizer.KEYWORDS,
312            "ANY TYPE": TokenType.VARIANT,
313            "BEGIN": TokenType.COMMAND,
314            "BEGIN TRANSACTION": TokenType.BEGIN,
315            "BYTES": TokenType.BINARY,
316            "CURRENT_DATETIME": TokenType.CURRENT_DATETIME,
317            "DATETIME": TokenType.TIMESTAMP,
318            "DECLARE": TokenType.COMMAND,
319            "ELSEIF": TokenType.COMMAND,
320            "EXCEPTION": TokenType.COMMAND,
321            "FLOAT64": TokenType.DOUBLE,
322            "FOR SYSTEM_TIME": TokenType.TIMESTAMP_SNAPSHOT,
323            "MODEL": TokenType.MODEL,
324            "NOT DETERMINISTIC": TokenType.VOLATILE,
325            "RECORD": TokenType.STRUCT,
326            "TIMESTAMP": TokenType.TIMESTAMPTZ,
327        }
328        KEYWORDS.pop("DIV")
329        KEYWORDS.pop("VALUES")
330
331    class Parser(parser.Parser):
332        PREFIXED_PIVOT_COLUMNS = True
333        LOG_DEFAULTS_TO_LN = True
334        SUPPORTS_IMPLICIT_UNNEST = True
335
336        FUNCTIONS = {
337            **parser.Parser.FUNCTIONS,
338            "DATE": _build_date,
339            "DATE_ADD": build_date_delta_with_interval(exp.DateAdd),
340            "DATE_SUB": build_date_delta_with_interval(exp.DateSub),
341            "DATE_TRUNC": lambda args: exp.DateTrunc(
342                unit=exp.Literal.string(str(seq_get(args, 1))),
343                this=seq_get(args, 0),
344            ),
345            "DATETIME_ADD": build_date_delta_with_interval(exp.DatetimeAdd),
346            "DATETIME_SUB": build_date_delta_with_interval(exp.DatetimeSub),
347            "DIV": binary_from_function(exp.IntDiv),
348            "FORMAT_DATE": lambda args: exp.TimeToStr(
349                this=exp.TsOrDsToDate(this=seq_get(args, 1)), format=seq_get(args, 0)
350            ),
351            "GENERATE_ARRAY": exp.GenerateSeries.from_arg_list,
352            "JSON_EXTRACT_SCALAR": lambda args: exp.JSONExtractScalar(
353                this=seq_get(args, 0), expression=seq_get(args, 1) or exp.Literal.string("$")
354            ),
355            "MD5": exp.MD5Digest.from_arg_list,
356            "TO_HEX": _build_to_hex,
357            "PARSE_DATE": lambda args: build_formatted_time(exp.StrToDate, "bigquery")(
358                [seq_get(args, 1), seq_get(args, 0)]
359            ),
360            "PARSE_TIMESTAMP": _build_parse_timestamp,
361            "REGEXP_CONTAINS": exp.RegexpLike.from_arg_list,
362            "REGEXP_EXTRACT": lambda args: exp.RegexpExtract(
363                this=seq_get(args, 0),
364                expression=seq_get(args, 1),
365                position=seq_get(args, 2),
366                occurrence=seq_get(args, 3),
367                group=exp.Literal.number(1) if re.compile(args[1].name).groups == 1 else None,
368            ),
369            "SHA256": lambda args: exp.SHA2(this=seq_get(args, 0), length=exp.Literal.number(256)),
370            "SHA512": lambda args: exp.SHA2(this=seq_get(args, 0), length=exp.Literal.number(512)),
371            "SPLIT": lambda args: exp.Split(
372                # https://cloud.google.com/bigquery/docs/reference/standard-sql/string_functions#split
373                this=seq_get(args, 0),
374                expression=seq_get(args, 1) or exp.Literal.string(","),
375            ),
376            "TIME": _build_time,
377            "TIME_ADD": build_date_delta_with_interval(exp.TimeAdd),
378            "TIME_SUB": build_date_delta_with_interval(exp.TimeSub),
379            "TIMESTAMP": _build_timestamp,
380            "TIMESTAMP_ADD": build_date_delta_with_interval(exp.TimestampAdd),
381            "TIMESTAMP_SUB": build_date_delta_with_interval(exp.TimestampSub),
382            "TIMESTAMP_MICROS": lambda args: exp.UnixToTime(
383                this=seq_get(args, 0), scale=exp.UnixToTime.MICROS
384            ),
385            "TIMESTAMP_MILLIS": lambda args: exp.UnixToTime(
386                this=seq_get(args, 0), scale=exp.UnixToTime.MILLIS
387            ),
388            "TIMESTAMP_SECONDS": lambda args: exp.UnixToTime(this=seq_get(args, 0)),
389            "TO_JSON_STRING": exp.JSONFormat.from_arg_list,
390        }
391
392        FUNCTION_PARSERS = {
393            **parser.Parser.FUNCTION_PARSERS,
394            "ARRAY": lambda self: self.expression(exp.Array, expressions=[self._parse_statement()]),
395        }
396        FUNCTION_PARSERS.pop("TRIM")
397
398        NO_PAREN_FUNCTIONS = {
399            **parser.Parser.NO_PAREN_FUNCTIONS,
400            TokenType.CURRENT_DATETIME: exp.CurrentDatetime,
401        }
402
403        NESTED_TYPE_TOKENS = {
404            *parser.Parser.NESTED_TYPE_TOKENS,
405            TokenType.TABLE,
406        }
407
408        PROPERTY_PARSERS = {
409            **parser.Parser.PROPERTY_PARSERS,
410            "NOT DETERMINISTIC": lambda self: self.expression(
411                exp.StabilityProperty, this=exp.Literal.string("VOLATILE")
412            ),
413            "OPTIONS": lambda self: self._parse_with_property(),
414        }
415
416        CONSTRAINT_PARSERS = {
417            **parser.Parser.CONSTRAINT_PARSERS,
418            "OPTIONS": lambda self: exp.Properties(expressions=self._parse_with_property()),
419        }
420
421        RANGE_PARSERS = parser.Parser.RANGE_PARSERS.copy()
422        RANGE_PARSERS.pop(TokenType.OVERLAPS)
423
424        NULL_TOKENS = {TokenType.NULL, TokenType.UNKNOWN}
425
426        STATEMENT_PARSERS = {
427            **parser.Parser.STATEMENT_PARSERS,
428            TokenType.ELSE: lambda self: self._parse_as_command(self._prev),
429            TokenType.END: lambda self: self._parse_as_command(self._prev),
430            TokenType.FOR: lambda self: self._parse_for_in(),
431        }
432
433        BRACKET_OFFSETS = {
434            "OFFSET": (0, False),
435            "ORDINAL": (1, False),
436            "SAFE_OFFSET": (0, True),
437            "SAFE_ORDINAL": (1, True),
438        }
439
440        def _parse_for_in(self) -> exp.ForIn:
441            this = self._parse_range()
442            self._match_text_seq("DO")
443            return self.expression(exp.ForIn, this=this, expression=self._parse_statement())
444
445        def _parse_table_part(self, schema: bool = False) -> t.Optional[exp.Expression]:
446            this = super()._parse_table_part(schema=schema) or self._parse_number()
447
448            # https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#table_names
449            if isinstance(this, exp.Identifier):
450                table_name = this.name
451                while self._match(TokenType.DASH, advance=False) and self._next:
452                    text = ""
453                    while self._curr and self._curr.token_type != TokenType.DOT:
454                        self._advance()
455                        text += self._prev.text
456                    table_name += text
457
458                this = exp.Identifier(this=table_name, quoted=this.args.get("quoted"))
459            elif isinstance(this, exp.Literal):
460                table_name = this.name
461
462                if self._is_connected() and self._parse_var(any_token=True):
463                    table_name += self._prev.text
464
465                this = exp.Identifier(this=table_name, quoted=True)
466
467            return this
468
469        def _parse_table_parts(
470            self, schema: bool = False, is_db_reference: bool = False, wildcard: bool = False
471        ) -> exp.Table:
472            table = super()._parse_table_parts(
473                schema=schema, is_db_reference=is_db_reference, wildcard=True
474            )
475
476            # proj-1.db.tbl -- `1.` is tokenized as a float so we need to unravel it here
477            if not table.catalog:
478                if table.db:
479                    parts = table.db.split(".")
480                    if len(parts) == 2 and not table.args["db"].quoted:
481                        table.set("catalog", exp.Identifier(this=parts[0]))
482                        table.set("db", exp.Identifier(this=parts[1]))
483                else:
484                    parts = table.name.split(".")
485                    if len(parts) == 2 and not table.this.quoted:
486                        table.set("db", exp.Identifier(this=parts[0]))
487                        table.set("this", exp.Identifier(this=parts[1]))
488
489            if any("." in p.name for p in table.parts):
490                catalog, db, this, *rest = (
491                    exp.to_identifier(p, quoted=True)
492                    for p in split_num_words(".".join(p.name for p in table.parts), ".", 3)
493                )
494
495                if rest and this:
496                    this = exp.Dot.build([this, *rest])  # type: ignore
497
498                table = exp.Table(this=this, db=db, catalog=catalog)
499                table.meta["quoted_table"] = True
500
501            return table
502
503        @t.overload
504        def _parse_json_object(self, agg: Lit[False]) -> exp.JSONObject: ...
505
506        @t.overload
507        def _parse_json_object(self, agg: Lit[True]) -> exp.JSONObjectAgg: ...
508
509        def _parse_json_object(self, agg=False):
510            json_object = super()._parse_json_object()
511            array_kv_pair = seq_get(json_object.expressions, 0)
512
513            # Converts BQ's "signature 2" of JSON_OBJECT into SQLGlot's canonical representation
514            # https://cloud.google.com/bigquery/docs/reference/standard-sql/json_functions#json_object_signature2
515            if (
516                array_kv_pair
517                and isinstance(array_kv_pair.this, exp.Array)
518                and isinstance(array_kv_pair.expression, exp.Array)
519            ):
520                keys = array_kv_pair.this.expressions
521                values = array_kv_pair.expression.expressions
522
523                json_object.set(
524                    "expressions",
525                    [exp.JSONKeyValue(this=k, expression=v) for k, v in zip(keys, values)],
526                )
527
528            return json_object
529
530        def _parse_bracket(
531            self, this: t.Optional[exp.Expression] = None
532        ) -> t.Optional[exp.Expression]:
533            bracket = super()._parse_bracket(this)
534
535            if this is bracket:
536                return bracket
537
538            if isinstance(bracket, exp.Bracket):
539                for expression in bracket.expressions:
540                    name = expression.name.upper()
541
542                    if name not in self.BRACKET_OFFSETS:
543                        break
544
545                    offset, safe = self.BRACKET_OFFSETS[name]
546                    bracket.set("offset", offset)
547                    bracket.set("safe", safe)
548                    expression.replace(expression.expressions[0])
549
550            return bracket
551
552    class Generator(generator.Generator):
553        EXPLICIT_UNION = True
554        INTERVAL_ALLOWS_PLURAL_FORM = False
555        JOIN_HINTS = False
556        QUERY_HINTS = False
557        TABLE_HINTS = False
558        LIMIT_FETCH = "LIMIT"
559        RENAME_TABLE_WITH_DB = False
560        NVL2_SUPPORTED = False
561        UNNEST_WITH_ORDINALITY = False
562        COLLATE_IS_FUNC = True
563        LIMIT_ONLY_LITERALS = True
564        SUPPORTS_TABLE_ALIAS_COLUMNS = False
565        UNPIVOT_ALIASES_ARE_IDENTIFIERS = False
566        JSON_KEY_VALUE_PAIR_SEP = ","
567        NULL_ORDERING_SUPPORTED = False
568        IGNORE_NULLS_IN_FUNC = True
569        JSON_PATH_SINGLE_QUOTE_ESCAPE = True
570        CAN_IMPLEMENT_ARRAY_ANY = True
571        SUPPORTS_TO_NUMBER = False
572        NAMED_PLACEHOLDER_TOKEN = "@"
573
574        TRANSFORMS = {
575            **generator.Generator.TRANSFORMS,
576            exp.ApproxDistinct: rename_func("APPROX_COUNT_DISTINCT"),
577            exp.ArgMax: arg_max_or_min_no_count("MAX_BY"),
578            exp.ArgMin: arg_max_or_min_no_count("MIN_BY"),
579            exp.ArrayContains: _array_contains_sql,
580            exp.ArrayFilter: filter_array_using_unnest,
581            exp.ArraySize: rename_func("ARRAY_LENGTH"),
582            exp.Cast: transforms.preprocess([transforms.remove_precision_parameterized_types]),
583            exp.CollateProperty: lambda self, e: (
584                f"DEFAULT COLLATE {self.sql(e, 'this')}"
585                if e.args.get("default")
586                else f"COLLATE {self.sql(e, 'this')}"
587            ),
588            exp.Commit: lambda *_: "COMMIT TRANSACTION",
589            exp.CountIf: rename_func("COUNTIF"),
590            exp.Create: _create_sql,
591            exp.CTE: transforms.preprocess([_pushdown_cte_column_names]),
592            exp.DateAdd: date_add_interval_sql("DATE", "ADD"),
593            exp.DateDiff: lambda self, e: self.func(
594                "DATE_DIFF", e.this, e.expression, e.unit or "DAY"
595            ),
596            exp.DateFromParts: rename_func("DATE"),
597            exp.DateStrToDate: datestrtodate_sql,
598            exp.DateSub: date_add_interval_sql("DATE", "SUB"),
599            exp.DatetimeAdd: date_add_interval_sql("DATETIME", "ADD"),
600            exp.DatetimeSub: date_add_interval_sql("DATETIME", "SUB"),
601            exp.DateTrunc: lambda self, e: self.func("DATE_TRUNC", e.this, e.text("unit")),
602            exp.FromTimeZone: lambda self, e: self.func(
603                "DATETIME", self.func("TIMESTAMP", e.this, e.args.get("zone")), "'UTC'"
604            ),
605            exp.GenerateSeries: rename_func("GENERATE_ARRAY"),
606            exp.GroupConcat: rename_func("STRING_AGG"),
607            exp.Hex: rename_func("TO_HEX"),
608            exp.If: if_sql(false_value="NULL"),
609            exp.ILike: no_ilike_sql,
610            exp.IntDiv: rename_func("DIV"),
611            exp.JSONFormat: rename_func("TO_JSON_STRING"),
612            exp.Max: max_or_greatest,
613            exp.Mod: rename_func("MOD"),
614            exp.MD5: lambda self, e: self.func("TO_HEX", self.func("MD5", e.this)),
615            exp.MD5Digest: rename_func("MD5"),
616            exp.Min: min_or_least,
617            exp.PartitionedByProperty: lambda self, e: f"PARTITION BY {self.sql(e, 'this')}",
618            exp.RegexpExtract: lambda self, e: self.func(
619                "REGEXP_EXTRACT",
620                e.this,
621                e.expression,
622                e.args.get("position"),
623                e.args.get("occurrence"),
624            ),
625            exp.RegexpReplace: regexp_replace_sql,
626            exp.RegexpLike: rename_func("REGEXP_CONTAINS"),
627            exp.ReturnsProperty: _returnsproperty_sql,
628            exp.Rollback: lambda *_: "ROLLBACK TRANSACTION",
629            exp.Select: transforms.preprocess(
630                [
631                    transforms.explode_to_unnest(),
632                    _unqualify_unnest,
633                    transforms.eliminate_distinct_on,
634                    _alias_ordered_group,
635                    transforms.eliminate_semi_and_anti_joins,
636                ]
637            ),
638            exp.SHA2: lambda self, e: self.func(
639                "SHA256" if e.text("length") == "256" else "SHA512", e.this
640            ),
641            exp.StabilityProperty: lambda self, e: (
642                "DETERMINISTIC" if e.name == "IMMUTABLE" else "NOT DETERMINISTIC"
643            ),
644            exp.StrToDate: lambda self, e: self.func("PARSE_DATE", self.format_time(e), e.this),
645            exp.StrToTime: lambda self, e: self.func(
646                "PARSE_TIMESTAMP", self.format_time(e), e.this, e.args.get("zone")
647            ),
648            exp.TimeAdd: date_add_interval_sql("TIME", "ADD"),
649            exp.TimeFromParts: rename_func("TIME"),
650            exp.TimeSub: date_add_interval_sql("TIME", "SUB"),
651            exp.TimestampAdd: date_add_interval_sql("TIMESTAMP", "ADD"),
652            exp.TimestampDiff: rename_func("TIMESTAMP_DIFF"),
653            exp.TimestampSub: date_add_interval_sql("TIMESTAMP", "SUB"),
654            exp.TimeStrToTime: timestrtotime_sql,
655            exp.Transaction: lambda *_: "BEGIN TRANSACTION",
656            exp.Trim: lambda self, e: self.func("TRIM", e.this, e.expression),
657            exp.TsOrDsAdd: _ts_or_ds_add_sql,
658            exp.TsOrDsDiff: _ts_or_ds_diff_sql,
659            exp.TsOrDsToTime: rename_func("TIME"),
660            exp.Unhex: rename_func("FROM_HEX"),
661            exp.UnixDate: rename_func("UNIX_DATE"),
662            exp.UnixToTime: _unix_to_time_sql,
663            exp.Values: _derived_table_values_to_unnest,
664            exp.VariancePop: rename_func("VAR_POP"),
665        }
666
667        SUPPORTED_JSON_PATH_PARTS = {
668            exp.JSONPathKey,
669            exp.JSONPathRoot,
670            exp.JSONPathSubscript,
671        }
672
673        TYPE_MAPPING = {
674            **generator.Generator.TYPE_MAPPING,
675            exp.DataType.Type.BIGDECIMAL: "BIGNUMERIC",
676            exp.DataType.Type.BIGINT: "INT64",
677            exp.DataType.Type.BINARY: "BYTES",
678            exp.DataType.Type.BOOLEAN: "BOOL",
679            exp.DataType.Type.CHAR: "STRING",
680            exp.DataType.Type.DECIMAL: "NUMERIC",
681            exp.DataType.Type.DOUBLE: "FLOAT64",
682            exp.DataType.Type.FLOAT: "FLOAT64",
683            exp.DataType.Type.INT: "INT64",
684            exp.DataType.Type.NCHAR: "STRING",
685            exp.DataType.Type.NVARCHAR: "STRING",
686            exp.DataType.Type.SMALLINT: "INT64",
687            exp.DataType.Type.TEXT: "STRING",
688            exp.DataType.Type.TIMESTAMP: "DATETIME",
689            exp.DataType.Type.TIMESTAMPTZ: "TIMESTAMP",
690            exp.DataType.Type.TIMESTAMPLTZ: "TIMESTAMP",
691            exp.DataType.Type.TINYINT: "INT64",
692            exp.DataType.Type.VARBINARY: "BYTES",
693            exp.DataType.Type.VARCHAR: "STRING",
694            exp.DataType.Type.VARIANT: "ANY TYPE",
695        }
696
697        PROPERTIES_LOCATION = {
698            **generator.Generator.PROPERTIES_LOCATION,
699            exp.PartitionedByProperty: exp.Properties.Location.POST_SCHEMA,
700            exp.VolatileProperty: exp.Properties.Location.UNSUPPORTED,
701        }
702
703        # from: https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#reserved_keywords
704        RESERVED_KEYWORDS = {
705            *generator.Generator.RESERVED_KEYWORDS,
706            "all",
707            "and",
708            "any",
709            "array",
710            "as",
711            "asc",
712            "assert_rows_modified",
713            "at",
714            "between",
715            "by",
716            "case",
717            "cast",
718            "collate",
719            "contains",
720            "create",
721            "cross",
722            "cube",
723            "current",
724            "default",
725            "define",
726            "desc",
727            "distinct",
728            "else",
729            "end",
730            "enum",
731            "escape",
732            "except",
733            "exclude",
734            "exists",
735            "extract",
736            "false",
737            "fetch",
738            "following",
739            "for",
740            "from",
741            "full",
742            "group",
743            "grouping",
744            "groups",
745            "hash",
746            "having",
747            "if",
748            "ignore",
749            "in",
750            "inner",
751            "intersect",
752            "interval",
753            "into",
754            "is",
755            "join",
756            "lateral",
757            "left",
758            "like",
759            "limit",
760            "lookup",
761            "merge",
762            "natural",
763            "new",
764            "no",
765            "not",
766            "null",
767            "nulls",
768            "of",
769            "on",
770            "or",
771            "order",
772            "outer",
773            "over",
774            "partition",
775            "preceding",
776            "proto",
777            "qualify",
778            "range",
779            "recursive",
780            "respect",
781            "right",
782            "rollup",
783            "rows",
784            "select",
785            "set",
786            "some",
787            "struct",
788            "tablesample",
789            "then",
790            "to",
791            "treat",
792            "true",
793            "unbounded",
794            "union",
795            "unnest",
796            "using",
797            "when",
798            "where",
799            "window",
800            "with",
801            "within",
802        }
803
804        def table_parts(self, expression: exp.Table) -> str:
805            # Depending on the context, `x.y` may not resolve to the same data source as `x`.`y`, so
806            # we need to make sure the correct quoting is used in each case.
807            #
808            # For example, if there is a CTE x that clashes with a schema name, then the former will
809            # return the table y in that schema, whereas the latter will return the CTE's y column:
810            #
811            # - WITH x AS (SELECT [1, 2] AS y) SELECT * FROM x, `x.y`   -> cross join
812            # - WITH x AS (SELECT [1, 2] AS y) SELECT * FROM x, `x`.`y` -> implicit unnest
813            if expression.meta.get("quoted_table"):
814                table_parts = ".".join(p.name for p in expression.parts)
815                return self.sql(exp.Identifier(this=table_parts, quoted=True))
816
817            return super().table_parts(expression)
818
819        def timetostr_sql(self, expression: exp.TimeToStr) -> str:
820            this = expression.this if isinstance(expression.this, exp.TsOrDsToDate) else expression
821            return self.func("FORMAT_DATE", self.format_time(expression), this.this)
822
823        def eq_sql(self, expression: exp.EQ) -> str:
824            # Operands of = cannot be NULL in BigQuery
825            if isinstance(expression.left, exp.Null) or isinstance(expression.right, exp.Null):
826                if not isinstance(expression.parent, exp.Update):
827                    return "NULL"
828
829            return self.binary(expression, "=")
830
831        def attimezone_sql(self, expression: exp.AtTimeZone) -> str:
832            parent = expression.parent
833
834            # BigQuery allows CAST(.. AS {STRING|TIMESTAMP} [FORMAT <fmt> [AT TIME ZONE <tz>]]).
835            # Only the TIMESTAMP one should use the below conversion, when AT TIME ZONE is included.
836            if not isinstance(parent, exp.Cast) or not parent.to.is_type("text"):
837                return self.func(
838                    "TIMESTAMP", self.func("DATETIME", expression.this, expression.args.get("zone"))
839                )
840
841            return super().attimezone_sql(expression)
842
843        def trycast_sql(self, expression: exp.TryCast) -> str:
844            return self.cast_sql(expression, safe_prefix="SAFE_")
845
846        def array_sql(self, expression: exp.Array) -> str:
847            first_arg = seq_get(expression.expressions, 0)
848            if isinstance(first_arg, exp.Query):
849                return f"ARRAY{self.wrap(self.sql(first_arg))}"
850
851            return inline_array_sql(self, expression)
852
853        def bracket_sql(self, expression: exp.Bracket) -> str:
854            this = expression.this
855            expressions = expression.expressions
856
857            if len(expressions) == 1 and this and this.is_type(exp.DataType.Type.STRUCT):
858                arg = expressions[0]
859                if arg.type is None:
860                    from sqlglot.optimizer.annotate_types import annotate_types
861
862                    arg = annotate_types(arg)
863
864                if arg.type and arg.type.this in exp.DataType.TEXT_TYPES:
865                    # BQ doesn't support bracket syntax with string values for structs
866                    return f"{self.sql(this)}.{arg.name}"
867
868            expressions_sql = self.expressions(expression, flat=True)
869            offset = expression.args.get("offset")
870
871            if offset == 0:
872                expressions_sql = f"OFFSET({expressions_sql})"
873            elif offset == 1:
874                expressions_sql = f"ORDINAL({expressions_sql})"
875            elif offset is not None:
876                self.unsupported(f"Unsupported array offset: {offset}")
877
878            if expression.args.get("safe"):
879                expressions_sql = f"SAFE_{expressions_sql}"
880
881            return f"{self.sql(this)}[{expressions_sql}]"
882
883        def in_unnest_op(self, expression: exp.Unnest) -> str:
884            return self.sql(expression)
885
886        def except_op(self, expression: exp.Except) -> str:
887            if not expression.args.get("distinct"):
888                self.unsupported("EXCEPT without DISTINCT is not supported in BigQuery")
889            return f"EXCEPT{' DISTINCT' if expression.args.get('distinct') else ' ALL'}"
890
891        def intersect_op(self, expression: exp.Intersect) -> str:
892            if not expression.args.get("distinct"):
893                self.unsupported("INTERSECT without DISTINCT is not supported in BigQuery")
894            return f"INTERSECT{' DISTINCT' if expression.args.get('distinct') else ' ALL'}"
895
896        def with_properties(self, properties: exp.Properties) -> str:
897            return self.properties(properties, prefix=self.seg("OPTIONS"))
898
899        def version_sql(self, expression: exp.Version) -> str:
900            if expression.name == "TIMESTAMP":
901                expression.set("this", "SYSTEM_TIME")
902            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)

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', '%E*S': '%S.%f', '%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] = {'_PARTITIONDATE', '_PARTITIONTIME'}

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:
270    def normalize_identifier(self, expression: E) -> E:
271        if isinstance(expression, exp.Identifier):
272            parent = expression.parent
273            while isinstance(parent, exp.Dot):
274                parent = parent.parent
275
276            # In BigQuery, CTEs are case-insensitive, but UDF and table names are case-sensitive
277            # by default. The following check uses a heuristic to detect tables based on whether
278            # they are qualified. This should generally be correct, because tables in BigQuery
279            # must be qualified with at least a dataset, unless @@dataset_id is set.
280            case_sensitive = (
281                isinstance(parent, exp.UserDefinedFunction)
282                or (
283                    isinstance(parent, exp.Table)
284                    and parent.db
285                    and (parent.meta.get("quoted_table") or not parent.meta.get("maybe_column"))
286                )
287                or expression.meta.get("is_table")
288            )
289            if not case_sensitive:
290                expression.set("this", expression.this.lower())
291
292        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.

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'>
parser_class = <class 'BigQuery.Parser'>
generator_class = <class 'BigQuery.Generator'>
TIME_TRIE: Dict = {'%': {'D': {0: True}, 'E': {'*': {'S': {0: True}}, '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):
294    class Tokenizer(tokens.Tokenizer):
295        QUOTES = ["'", '"', '"""', "'''"]
296        COMMENTS = ["--", "#", ("/*", "*/")]
297        IDENTIFIERS = ["`"]
298        STRING_ESCAPES = ["\\"]
299
300        HEX_STRINGS = [("0x", ""), ("0X", "")]
301
302        BYTE_STRINGS = [
303            (prefix + q, q) for q in t.cast(t.List[str], QUOTES) for prefix in ("b", "B")
304        ]
305
306        RAW_STRINGS = [
307            (prefix + q, q) for q in t.cast(t.List[str], QUOTES) for prefix in ("r", "R")
308        ]
309
310        KEYWORDS = {
311            **tokens.Tokenizer.KEYWORDS,
312            "ANY TYPE": TokenType.VARIANT,
313            "BEGIN": TokenType.COMMAND,
314            "BEGIN TRANSACTION": TokenType.BEGIN,
315            "BYTES": TokenType.BINARY,
316            "CURRENT_DATETIME": TokenType.CURRENT_DATETIME,
317            "DATETIME": TokenType.TIMESTAMP,
318            "DECLARE": TokenType.COMMAND,
319            "ELSEIF": TokenType.COMMAND,
320            "EXCEPTION": TokenType.COMMAND,
321            "FLOAT64": TokenType.DOUBLE,
322            "FOR SYSTEM_TIME": TokenType.TIMESTAMP_SNAPSHOT,
323            "MODEL": TokenType.MODEL,
324            "NOT DETERMINISTIC": TokenType.VOLATILE,
325            "RECORD": TokenType.STRUCT,
326            "TIMESTAMP": TokenType.TIMESTAMPTZ,
327        }
328        KEYWORDS.pop("DIV")
329        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'>, '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'>, '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'>, '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'>, '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'>, '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'>, 'COPY': <TokenType.COMMAND: 'COMMAND'>, '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):
331    class Parser(parser.Parser):
332        PREFIXED_PIVOT_COLUMNS = True
333        LOG_DEFAULTS_TO_LN = True
334        SUPPORTS_IMPLICIT_UNNEST = True
335
336        FUNCTIONS = {
337            **parser.Parser.FUNCTIONS,
338            "DATE": _build_date,
339            "DATE_ADD": build_date_delta_with_interval(exp.DateAdd),
340            "DATE_SUB": build_date_delta_with_interval(exp.DateSub),
341            "DATE_TRUNC": lambda args: exp.DateTrunc(
342                unit=exp.Literal.string(str(seq_get(args, 1))),
343                this=seq_get(args, 0),
344            ),
345            "DATETIME_ADD": build_date_delta_with_interval(exp.DatetimeAdd),
346            "DATETIME_SUB": build_date_delta_with_interval(exp.DatetimeSub),
347            "DIV": binary_from_function(exp.IntDiv),
348            "FORMAT_DATE": lambda args: exp.TimeToStr(
349                this=exp.TsOrDsToDate(this=seq_get(args, 1)), format=seq_get(args, 0)
350            ),
351            "GENERATE_ARRAY": exp.GenerateSeries.from_arg_list,
352            "JSON_EXTRACT_SCALAR": lambda args: exp.JSONExtractScalar(
353                this=seq_get(args, 0), expression=seq_get(args, 1) or exp.Literal.string("$")
354            ),
355            "MD5": exp.MD5Digest.from_arg_list,
356            "TO_HEX": _build_to_hex,
357            "PARSE_DATE": lambda args: build_formatted_time(exp.StrToDate, "bigquery")(
358                [seq_get(args, 1), seq_get(args, 0)]
359            ),
360            "PARSE_TIMESTAMP": _build_parse_timestamp,
361            "REGEXP_CONTAINS": exp.RegexpLike.from_arg_list,
362            "REGEXP_EXTRACT": lambda args: exp.RegexpExtract(
363                this=seq_get(args, 0),
364                expression=seq_get(args, 1),
365                position=seq_get(args, 2),
366                occurrence=seq_get(args, 3),
367                group=exp.Literal.number(1) if re.compile(args[1].name).groups == 1 else None,
368            ),
369            "SHA256": lambda args: exp.SHA2(this=seq_get(args, 0), length=exp.Literal.number(256)),
370            "SHA512": lambda args: exp.SHA2(this=seq_get(args, 0), length=exp.Literal.number(512)),
371            "SPLIT": lambda args: exp.Split(
372                # https://cloud.google.com/bigquery/docs/reference/standard-sql/string_functions#split
373                this=seq_get(args, 0),
374                expression=seq_get(args, 1) or exp.Literal.string(","),
375            ),
376            "TIME": _build_time,
377            "TIME_ADD": build_date_delta_with_interval(exp.TimeAdd),
378            "TIME_SUB": build_date_delta_with_interval(exp.TimeSub),
379            "TIMESTAMP": _build_timestamp,
380            "TIMESTAMP_ADD": build_date_delta_with_interval(exp.TimestampAdd),
381            "TIMESTAMP_SUB": build_date_delta_with_interval(exp.TimestampSub),
382            "TIMESTAMP_MICROS": lambda args: exp.UnixToTime(
383                this=seq_get(args, 0), scale=exp.UnixToTime.MICROS
384            ),
385            "TIMESTAMP_MILLIS": lambda args: exp.UnixToTime(
386                this=seq_get(args, 0), scale=exp.UnixToTime.MILLIS
387            ),
388            "TIMESTAMP_SECONDS": lambda args: exp.UnixToTime(this=seq_get(args, 0)),
389            "TO_JSON_STRING": exp.JSONFormat.from_arg_list,
390        }
391
392        FUNCTION_PARSERS = {
393            **parser.Parser.FUNCTION_PARSERS,
394            "ARRAY": lambda self: self.expression(exp.Array, expressions=[self._parse_statement()]),
395        }
396        FUNCTION_PARSERS.pop("TRIM")
397
398        NO_PAREN_FUNCTIONS = {
399            **parser.Parser.NO_PAREN_FUNCTIONS,
400            TokenType.CURRENT_DATETIME: exp.CurrentDatetime,
401        }
402
403        NESTED_TYPE_TOKENS = {
404            *parser.Parser.NESTED_TYPE_TOKENS,
405            TokenType.TABLE,
406        }
407
408        PROPERTY_PARSERS = {
409            **parser.Parser.PROPERTY_PARSERS,
410            "NOT DETERMINISTIC": lambda self: self.expression(
411                exp.StabilityProperty, this=exp.Literal.string("VOLATILE")
412            ),
413            "OPTIONS": lambda self: self._parse_with_property(),
414        }
415
416        CONSTRAINT_PARSERS = {
417            **parser.Parser.CONSTRAINT_PARSERS,
418            "OPTIONS": lambda self: exp.Properties(expressions=self._parse_with_property()),
419        }
420
421        RANGE_PARSERS = parser.Parser.RANGE_PARSERS.copy()
422        RANGE_PARSERS.pop(TokenType.OVERLAPS)
423
424        NULL_TOKENS = {TokenType.NULL, TokenType.UNKNOWN}
425
426        STATEMENT_PARSERS = {
427            **parser.Parser.STATEMENT_PARSERS,
428            TokenType.ELSE: lambda self: self._parse_as_command(self._prev),
429            TokenType.END: lambda self: self._parse_as_command(self._prev),
430            TokenType.FOR: lambda self: self._parse_for_in(),
431        }
432
433        BRACKET_OFFSETS = {
434            "OFFSET": (0, False),
435            "ORDINAL": (1, False),
436            "SAFE_OFFSET": (0, True),
437            "SAFE_ORDINAL": (1, True),
438        }
439
440        def _parse_for_in(self) -> exp.ForIn:
441            this = self._parse_range()
442            self._match_text_seq("DO")
443            return self.expression(exp.ForIn, this=this, expression=self._parse_statement())
444
445        def _parse_table_part(self, schema: bool = False) -> t.Optional[exp.Expression]:
446            this = super()._parse_table_part(schema=schema) or self._parse_number()
447
448            # https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#table_names
449            if isinstance(this, exp.Identifier):
450                table_name = this.name
451                while self._match(TokenType.DASH, advance=False) and self._next:
452                    text = ""
453                    while self._curr and self._curr.token_type != TokenType.DOT:
454                        self._advance()
455                        text += self._prev.text
456                    table_name += text
457
458                this = exp.Identifier(this=table_name, quoted=this.args.get("quoted"))
459            elif isinstance(this, exp.Literal):
460                table_name = this.name
461
462                if self._is_connected() and self._parse_var(any_token=True):
463                    table_name += self._prev.text
464
465                this = exp.Identifier(this=table_name, quoted=True)
466
467            return this
468
469        def _parse_table_parts(
470            self, schema: bool = False, is_db_reference: bool = False, wildcard: bool = False
471        ) -> exp.Table:
472            table = super()._parse_table_parts(
473                schema=schema, is_db_reference=is_db_reference, wildcard=True
474            )
475
476            # proj-1.db.tbl -- `1.` is tokenized as a float so we need to unravel it here
477            if not table.catalog:
478                if table.db:
479                    parts = table.db.split(".")
480                    if len(parts) == 2 and not table.args["db"].quoted:
481                        table.set("catalog", exp.Identifier(this=parts[0]))
482                        table.set("db", exp.Identifier(this=parts[1]))
483                else:
484                    parts = table.name.split(".")
485                    if len(parts) == 2 and not table.this.quoted:
486                        table.set("db", exp.Identifier(this=parts[0]))
487                        table.set("this", exp.Identifier(this=parts[1]))
488
489            if any("." in p.name for p in table.parts):
490                catalog, db, this, *rest = (
491                    exp.to_identifier(p, quoted=True)
492                    for p in split_num_words(".".join(p.name for p in table.parts), ".", 3)
493                )
494
495                if rest and this:
496                    this = exp.Dot.build([this, *rest])  # type: ignore
497
498                table = exp.Table(this=this, db=db, catalog=catalog)
499                table.meta["quoted_table"] = True
500
501            return table
502
503        @t.overload
504        def _parse_json_object(self, agg: Lit[False]) -> exp.JSONObject: ...
505
506        @t.overload
507        def _parse_json_object(self, agg: Lit[True]) -> exp.JSONObjectAgg: ...
508
509        def _parse_json_object(self, agg=False):
510            json_object = super()._parse_json_object()
511            array_kv_pair = seq_get(json_object.expressions, 0)
512
513            # Converts BQ's "signature 2" of JSON_OBJECT into SQLGlot's canonical representation
514            # https://cloud.google.com/bigquery/docs/reference/standard-sql/json_functions#json_object_signature2
515            if (
516                array_kv_pair
517                and isinstance(array_kv_pair.this, exp.Array)
518                and isinstance(array_kv_pair.expression, exp.Array)
519            ):
520                keys = array_kv_pair.this.expressions
521                values = array_kv_pair.expression.expressions
522
523                json_object.set(
524                    "expressions",
525                    [exp.JSONKeyValue(this=k, expression=v) for k, v in zip(keys, values)],
526                )
527
528            return json_object
529
530        def _parse_bracket(
531            self, this: t.Optional[exp.Expression] = None
532        ) -> t.Optional[exp.Expression]:
533            bracket = super()._parse_bracket(this)
534
535            if this is bracket:
536                return bracket
537
538            if isinstance(bracket, exp.Bracket):
539                for expression in bracket.expressions:
540                    name = expression.name.upper()
541
542                    if name not in self.BRACKET_OFFSETS:
543                        break
544
545                    offset, safe = self.BRACKET_OFFSETS[name]
546                    bracket.set("offset", offset)
547                    bracket.set("safe", safe)
548                    expression.replace(expression.expressions[0])
549
550            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_CONTAINS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayContains'>>, '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_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'>>, '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': <bound method Func.from_arg_list of <class 'sqlglot.expressions.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_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'>>, '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': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Lower'>>, 'LCASE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Lower'>>, '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'>>, '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'>>, '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_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_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'>>, 'UPPER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.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 Parser.<lambda>>, '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'>>, 'TO_HEX': <function _build_to_hex>, '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>>, 'TIME': <function _build_time>, '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>>, '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.STRUCT: 'STRUCT'>, <TokenType.TABLE: 'TABLE'>, <TokenType.MAP: 'MAP'>, <TokenType.NESTED: 'NESTED'>, <TokenType.LOWCARDINALITY: 'LOWCARDINALITY'>, <TokenType.ARRAY: 'ARRAY'>, <TokenType.OBJECT: 'OBJECT'>, <TokenType.NULLABLE: 'NULLABLE'>}
PROPERTY_PARSERS = {'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>>, '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>>, 'ROW': <function Parser.<lambda>>, 'ROW_FORMAT': <function Parser.<lambda>>, 'SAMPLE': <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>>, '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>.<lambda>>, <TokenType.ILIKE: 'ILIKE'>: <function binary_range_parser.<locals>.<lambda>>, <TokenType.IN: 'IN'>: <function Parser.<lambda>>, <TokenType.IRLIKE: 'IRLIKE'>: <function binary_range_parser.<locals>.<lambda>>, <TokenType.IS: 'IS'>: <function Parser.<lambda>>, <TokenType.LIKE: 'LIKE'>: <function binary_range_parser.<locals>.<lambda>>, <TokenType.RLIKE: 'RLIKE'>: <function binary_range_parser.<locals>.<lambda>>, <TokenType.SIMILAR_TO: 'SIMILAR_TO'>: <function binary_range_parser.<locals>.<lambda>>, <TokenType.FOR: 'FOR'>: <function Parser.<lambda>>}
NULL_TOKENS = {<TokenType.NULL: 'NULL'>, <TokenType.UNKNOWN: 'UNKNOWN'>}
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.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.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)}
TABLE_ALIAS_TOKENS = {<TokenType.PERCENT: 'PERCENT'>, <TokenType.ALL: 'ALL'>, <TokenType.DELETE: 'DELETE'>, <TokenType.CURRENT_TIME: 'CURRENT_TIME'>, <TokenType.ROWVERSION: 'ROWVERSION'>, <TokenType.TRUNCATE: 'TRUNCATE'>, <TokenType.NULLABLE: 'NULLABLE'>, <TokenType.VAR: 'VAR'>, <TokenType.TEMPORARY: 'TEMPORARY'>, <TokenType.UNIQUEIDENTIFIER: 'UNIQUEIDENTIFIER'>, <TokenType.DATETIME: 'DATETIME'>, <TokenType.EXISTS: 'EXISTS'>, <TokenType.INET: 'INET'>, <TokenType.INTERVAL: 'INTERVAL'>, <TokenType.SERIAL: 'SERIAL'>, <TokenType.TIMESTAMPTZ: 'TIMESTAMPTZ'>, <TokenType.ANY: 'ANY'>, <TokenType.IPV4: 'IPV4'>, <TokenType.FUNCTION: 'FUNCTION'>, <TokenType.TIMESTAMP_NS: 'TIMESTAMP_NS'>, <TokenType.FILTER: 'FILTER'>, <TokenType.SEQUENCE: 'SEQUENCE'>, <TokenType.TABLE: 'TABLE'>, <TokenType.NESTED: 'NESTED'>, <TokenType.YEAR: 'YEAR'>, <TokenType.DATETIME64: 'DATETIME64'>, <TokenType.NEXT: 'NEXT'>, <TokenType.MODEL: 'MODEL'>, <TokenType.TIMESTAMP: 'TIMESTAMP'>, <TokenType.OBJECT_IDENTIFIER: 'OBJECT_IDENTIFIER'>, <TokenType.INDEX: 'INDEX'>, <TokenType.INT4RANGE: 'INT4RANGE'>, <TokenType.FALSE: 'FALSE'>, <TokenType.NAME: 'NAME'>, <TokenType.NULL: 'NULL'>, <TokenType.VARBINARY: 'VARBINARY'>, <TokenType.TEXT: 'TEXT'>, <TokenType.FIRST: 'FIRST'>, <TokenType.TIMESTAMP_S: 'TIMESTAMP_S'>, <TokenType.AUTO_INCREMENT: 'AUTO_INCREMENT'>, <TokenType.LONGBLOB: 'LONGBLOB'>, <TokenType.SMALLSERIAL: 'SMALLSERIAL'>, <TokenType.FLOAT: 'FLOAT'>, <TokenType.INT256: 'INT256'>, <TokenType.RANGE: 'RANGE'>, <TokenType.FOREIGN_KEY: 'FOREIGN_KEY'>, <TokenType.BEGIN: 'BEGIN'>, <TokenType.DATE32: 'DATE32'>, <TokenType.COMMAND: 'COMMAND'>, <TokenType.ISNULL: 'ISNULL'>, <TokenType.SCHEMA: 'SCHEMA'>, <TokenType.SMALLINT: 'SMALLINT'>, <TokenType.REFERENCES: 'REFERENCES'>, <TokenType.UNIQUE: 'UNIQUE'>, <TokenType.KILL: 'KILL'>, <TokenType.ORDINALITY: 'ORDINALITY'>, <TokenType.UDECIMAL: 'UDECIMAL'>, <TokenType.CURRENT_DATE: 'CURRENT_DATE'>, <TokenType.TSRANGE: 'TSRANGE'>, <TokenType.UMEDIUMINT: 'UMEDIUMINT'>, <TokenType.TIME: 'TIME'>, <TokenType.OBJECT: 'OBJECT'>, <TokenType.OPERATOR: 'OPERATOR'>, <TokenType.LOAD: 'LOAD'>, <TokenType.MERGE: 'MERGE'>, <TokenType.BPCHAR: 'BPCHAR'>, <TokenType.BIGINT: 'BIGINT'>, <TokenType.UNPIVOT: 'UNPIVOT'>, <TokenType.ENUM8: 'ENUM8'>, <TokenType.CURRENT_DATETIME: 'CURRENT_DATETIME'>, <TokenType.COLUMN: 'COLUMN'>, <TokenType.JSON: 'JSON'>, <TokenType.IPV6: 'IPV6'>, <TokenType.CONSTRAINT: 'CONSTRAINT'>, <TokenType.UINT128: 'UINT128'>, <TokenType.VOLATILE: 'VOLATILE'>, <TokenType.SOME: 'SOME'>, <TokenType.BOOLEAN: 'BOOLEAN'>, <TokenType.MAP: 'MAP'>, <TokenType.UPDATE: 'UPDATE'>, <TokenType.CASE: 'CASE'>, <TokenType.ROW: 'ROW'>, <TokenType.TOP: 'TOP'>, <TokenType.DECIMAL: 'DECIMAL'>, <TokenType.PIVOT: 'PIVOT'>, <TokenType.LOWCARDINALITY: 'LOWCARDINALITY'>, <TokenType.NVARCHAR: 'NVARCHAR'>, <TokenType.TINYBLOB: 'TINYBLOB'>, <TokenType.UINT: 'UINT'>, <TokenType.MEDIUMTEXT: 'MEDIUMTEXT'>, <TokenType.ARRAY: 'ARRAY'>, <TokenType.ASC: 'ASC'>, <TokenType.INT128: 'INT128'>, <TokenType.TRUE: 'TRUE'>, <TokenType.TSMULTIRANGE: 'TSMULTIRANGE'>, <TokenType.NCHAR: 'NCHAR'>, <TokenType.SEMI: 'SEMI'>, <TokenType.VARCHAR: 'VARCHAR'>, <TokenType.BIGSERIAL: 'BIGSERIAL'>, <TokenType.HLLSKETCH: 'HLLSKETCH'>, <TokenType.XML: 'XML'>, <TokenType.PARTITION: 'PARTITION'>, <TokenType.UNKNOWN: 'UNKNOWN'>, <TokenType.OVERWRITE: 'OVERWRITE'>, <TokenType.DATEMULTIRANGE: 'DATEMULTIRANGE'>, <TokenType.TIMESTAMP_MS: 'TIMESTAMP_MS'>, <TokenType.DATERANGE: 'DATERANGE'>, <TokenType.BIGDECIMAL: 'BIGDECIMAL'>, <TokenType.CURRENT_USER: 'CURRENT_USER'>, <TokenType.BINARY: 'BINARY'>, <TokenType.LONGTEXT: 'LONGTEXT'>, <TokenType.STORAGE_INTEGRATION: 'STORAGE_INTEGRATION'>, <TokenType.DATE: 'DATE'>, <TokenType.INT4MULTIRANGE: 'INT4MULTIRANGE'>, <TokenType.MEDIUMINT: 'MEDIUMINT'>, <TokenType.GEOGRAPHY: 'GEOGRAPHY'>, <TokenType.REPLACE: 'REPLACE'>, <TokenType.IPADDRESS: 'IPADDRESS'>, <TokenType.TIMESTAMPLTZ: 'TIMESTAMPLTZ'>, <TokenType.HSTORE: 'HSTORE'>, <TokenType.CURRENT_TIMESTAMP: 'CURRENT_TIMESTAMP'>, <TokenType.FORMAT: 'FORMAT'>, <TokenType.SIMPLEAGGREGATEFUNCTION: 'SIMPLEAGGREGATEFUNCTION'>, <TokenType.REFRESH: 'REFRESH'>, <TokenType.BIT: 'BIT'>, <TokenType.FINAL: 'FINAL'>, <TokenType.SHOW: 'SHOW'>, <TokenType.MONEY: 'MONEY'>, <TokenType.GEOMETRY: 'GEOMETRY'>, <TokenType.USE: 'USE'>, <TokenType.IMAGE: 'IMAGE'>, <TokenType.PROCEDURE: 'PROCEDURE'>, <TokenType.AGGREGATEFUNCTION: 'AGGREGATEFUNCTION'>, <TokenType.PSEUDO_TYPE: 'PSEUDO_TYPE'>, <TokenType.ESCAPE: 'ESCAPE'>, <TokenType.DICTIONARY: 'DICTIONARY'>, <TokenType.CACHE: 'CACHE'>, <TokenType.FIXEDSTRING: 'FIXEDSTRING'>, <TokenType.UINT256: 'UINT256'>, <TokenType.UBIGINT: 'UBIGINT'>, <TokenType.DATABASE: 'DATABASE'>, <TokenType.TINYTEXT: 'TINYTEXT'>, <TokenType.ENUM: 'ENUM'>, <TokenType.VIEW: 'VIEW'>, <TokenType.SETTINGS: 'SETTINGS'>, <TokenType.NUMRANGE: 'NUMRANGE'>, <TokenType.IPPREFIX: 'IPPREFIX'>, <TokenType.IS: 'IS'>, <TokenType.USERDEFINED: 'USERDEFINED'>, <TokenType.ENUM16: 'ENUM16'>, <TokenType.STRUCT: 'STRUCT'>, <TokenType.DIV: 'DIV'>, <TokenType.TSTZMULTIRANGE: 'TSTZMULTIRANGE'>, <TokenType.SUPER: 'SUPER'>, <TokenType.SMALLMONEY: 'SMALLMONEY'>, <TokenType.DESC: 'DESC'>, <TokenType.DESCRIBE: 'DESCRIBE'>, <TokenType.COMMIT: 'COMMIT'>, <TokenType.SET: 'SET'>, <TokenType.CHAR: 'CHAR'>, <TokenType.UTINYINT: 'UTINYINT'>, <TokenType.COMMENT: 'COMMENT'>, <TokenType.INT8RANGE: 'INT8RANGE'>, <TokenType.NUMMULTIRANGE: 'NUMMULTIRANGE'>, <TokenType.PRAGMA: 'PRAGMA'>, <TokenType.INT: 'INT'>, <TokenType.USMALLINT: 'USMALLINT'>, <TokenType.DOUBLE: 'DOUBLE'>, <TokenType.KEEP: 'KEEP'>, <TokenType.ANTI: 'ANTI'>, <TokenType.EXECUTE: 'EXECUTE'>, <TokenType.END: 'END'>, <TokenType.UUID: 'UUID'>, <TokenType.VARIANT: 'VARIANT'>, <TokenType.MEDIUMBLOB: 'MEDIUMBLOB'>, <TokenType.INT8MULTIRANGE: 'INT8MULTIRANGE'>, <TokenType.TIMETZ: 'TIMETZ'>, <TokenType.OVERLAPS: 'OVERLAPS'>, <TokenType.ROWS: 'ROWS'>, <TokenType.COLLATE: 'COLLATE'>, <TokenType.RECURSIVE: 'RECURSIVE'>, <TokenType.JSONB: 'JSONB'>, <TokenType.DEFAULT: 'DEFAULT'>, <TokenType.TSTZRANGE: 'TSTZRANGE'>, <TokenType.TINYINT: 'TINYINT'>}
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
ID_VAR_TOKENS
INTERVAL_VARS
ALIAS_TOKENS
COMMENT_TABLE_ALIAS_TOKENS
UPDATE_ALIAS_TOKENS
TRIM_TYPES
FUNC_TOKENS
CONJUNCTION
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
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
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
STRICT_CAST
IDENTIFY_PIVOT_STRINGS
ALTER_TABLE_ADD_REQUIRED_FOR_EACH_COLUMN
TABLESAMPLE_CSV
SET_REQUIRES_ASSIGNMENT_DELIMITER
TRIM_PATTERN_FIRST
STRING_ALIASES
MODIFIERS_ATTACHED_TO_UNION
UNION_MODIFIERS
NO_PAREN_IF_COMMANDS
JSON_ARROWS_REQUIRE_JSON_TYPE
VALUES_FOLLOWED_BY_PAREN
INTERVAL_SPANS
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):
552    class Generator(generator.Generator):
553        EXPLICIT_UNION = True
554        INTERVAL_ALLOWS_PLURAL_FORM = False
555        JOIN_HINTS = False
556        QUERY_HINTS = False
557        TABLE_HINTS = False
558        LIMIT_FETCH = "LIMIT"
559        RENAME_TABLE_WITH_DB = False
560        NVL2_SUPPORTED = False
561        UNNEST_WITH_ORDINALITY = False
562        COLLATE_IS_FUNC = True
563        LIMIT_ONLY_LITERALS = True
564        SUPPORTS_TABLE_ALIAS_COLUMNS = False
565        UNPIVOT_ALIASES_ARE_IDENTIFIERS = False
566        JSON_KEY_VALUE_PAIR_SEP = ","
567        NULL_ORDERING_SUPPORTED = False
568        IGNORE_NULLS_IN_FUNC = True
569        JSON_PATH_SINGLE_QUOTE_ESCAPE = True
570        CAN_IMPLEMENT_ARRAY_ANY = True
571        SUPPORTS_TO_NUMBER = False
572        NAMED_PLACEHOLDER_TOKEN = "@"
573
574        TRANSFORMS = {
575            **generator.Generator.TRANSFORMS,
576            exp.ApproxDistinct: rename_func("APPROX_COUNT_DISTINCT"),
577            exp.ArgMax: arg_max_or_min_no_count("MAX_BY"),
578            exp.ArgMin: arg_max_or_min_no_count("MIN_BY"),
579            exp.ArrayContains: _array_contains_sql,
580            exp.ArrayFilter: filter_array_using_unnest,
581            exp.ArraySize: rename_func("ARRAY_LENGTH"),
582            exp.Cast: transforms.preprocess([transforms.remove_precision_parameterized_types]),
583            exp.CollateProperty: lambda self, e: (
584                f"DEFAULT COLLATE {self.sql(e, 'this')}"
585                if e.args.get("default")
586                else f"COLLATE {self.sql(e, 'this')}"
587            ),
588            exp.Commit: lambda *_: "COMMIT TRANSACTION",
589            exp.CountIf: rename_func("COUNTIF"),
590            exp.Create: _create_sql,
591            exp.CTE: transforms.preprocess([_pushdown_cte_column_names]),
592            exp.DateAdd: date_add_interval_sql("DATE", "ADD"),
593            exp.DateDiff: lambda self, e: self.func(
594                "DATE_DIFF", e.this, e.expression, e.unit or "DAY"
595            ),
596            exp.DateFromParts: rename_func("DATE"),
597            exp.DateStrToDate: datestrtodate_sql,
598            exp.DateSub: date_add_interval_sql("DATE", "SUB"),
599            exp.DatetimeAdd: date_add_interval_sql("DATETIME", "ADD"),
600            exp.DatetimeSub: date_add_interval_sql("DATETIME", "SUB"),
601            exp.DateTrunc: lambda self, e: self.func("DATE_TRUNC", e.this, e.text("unit")),
602            exp.FromTimeZone: lambda self, e: self.func(
603                "DATETIME", self.func("TIMESTAMP", e.this, e.args.get("zone")), "'UTC'"
604            ),
605            exp.GenerateSeries: rename_func("GENERATE_ARRAY"),
606            exp.GroupConcat: rename_func("STRING_AGG"),
607            exp.Hex: rename_func("TO_HEX"),
608            exp.If: if_sql(false_value="NULL"),
609            exp.ILike: no_ilike_sql,
610            exp.IntDiv: rename_func("DIV"),
611            exp.JSONFormat: rename_func("TO_JSON_STRING"),
612            exp.Max: max_or_greatest,
613            exp.Mod: rename_func("MOD"),
614            exp.MD5: lambda self, e: self.func("TO_HEX", self.func("MD5", e.this)),
615            exp.MD5Digest: rename_func("MD5"),
616            exp.Min: min_or_least,
617            exp.PartitionedByProperty: lambda self, e: f"PARTITION BY {self.sql(e, 'this')}",
618            exp.RegexpExtract: lambda self, e: self.func(
619                "REGEXP_EXTRACT",
620                e.this,
621                e.expression,
622                e.args.get("position"),
623                e.args.get("occurrence"),
624            ),
625            exp.RegexpReplace: regexp_replace_sql,
626            exp.RegexpLike: rename_func("REGEXP_CONTAINS"),
627            exp.ReturnsProperty: _returnsproperty_sql,
628            exp.Rollback: lambda *_: "ROLLBACK TRANSACTION",
629            exp.Select: transforms.preprocess(
630                [
631                    transforms.explode_to_unnest(),
632                    _unqualify_unnest,
633                    transforms.eliminate_distinct_on,
634                    _alias_ordered_group,
635                    transforms.eliminate_semi_and_anti_joins,
636                ]
637            ),
638            exp.SHA2: lambda self, e: self.func(
639                "SHA256" if e.text("length") == "256" else "SHA512", e.this
640            ),
641            exp.StabilityProperty: lambda self, e: (
642                "DETERMINISTIC" if e.name == "IMMUTABLE" else "NOT DETERMINISTIC"
643            ),
644            exp.StrToDate: lambda self, e: self.func("PARSE_DATE", self.format_time(e), e.this),
645            exp.StrToTime: lambda self, e: self.func(
646                "PARSE_TIMESTAMP", self.format_time(e), e.this, e.args.get("zone")
647            ),
648            exp.TimeAdd: date_add_interval_sql("TIME", "ADD"),
649            exp.TimeFromParts: rename_func("TIME"),
650            exp.TimeSub: date_add_interval_sql("TIME", "SUB"),
651            exp.TimestampAdd: date_add_interval_sql("TIMESTAMP", "ADD"),
652            exp.TimestampDiff: rename_func("TIMESTAMP_DIFF"),
653            exp.TimestampSub: date_add_interval_sql("TIMESTAMP", "SUB"),
654            exp.TimeStrToTime: timestrtotime_sql,
655            exp.Transaction: lambda *_: "BEGIN TRANSACTION",
656            exp.Trim: lambda self, e: self.func("TRIM", e.this, e.expression),
657            exp.TsOrDsAdd: _ts_or_ds_add_sql,
658            exp.TsOrDsDiff: _ts_or_ds_diff_sql,
659            exp.TsOrDsToTime: rename_func("TIME"),
660            exp.Unhex: rename_func("FROM_HEX"),
661            exp.UnixDate: rename_func("UNIX_DATE"),
662            exp.UnixToTime: _unix_to_time_sql,
663            exp.Values: _derived_table_values_to_unnest,
664            exp.VariancePop: rename_func("VAR_POP"),
665        }
666
667        SUPPORTED_JSON_PATH_PARTS = {
668            exp.JSONPathKey,
669            exp.JSONPathRoot,
670            exp.JSONPathSubscript,
671        }
672
673        TYPE_MAPPING = {
674            **generator.Generator.TYPE_MAPPING,
675            exp.DataType.Type.BIGDECIMAL: "BIGNUMERIC",
676            exp.DataType.Type.BIGINT: "INT64",
677            exp.DataType.Type.BINARY: "BYTES",
678            exp.DataType.Type.BOOLEAN: "BOOL",
679            exp.DataType.Type.CHAR: "STRING",
680            exp.DataType.Type.DECIMAL: "NUMERIC",
681            exp.DataType.Type.DOUBLE: "FLOAT64",
682            exp.DataType.Type.FLOAT: "FLOAT64",
683            exp.DataType.Type.INT: "INT64",
684            exp.DataType.Type.NCHAR: "STRING",
685            exp.DataType.Type.NVARCHAR: "STRING",
686            exp.DataType.Type.SMALLINT: "INT64",
687            exp.DataType.Type.TEXT: "STRING",
688            exp.DataType.Type.TIMESTAMP: "DATETIME",
689            exp.DataType.Type.TIMESTAMPTZ: "TIMESTAMP",
690            exp.DataType.Type.TIMESTAMPLTZ: "TIMESTAMP",
691            exp.DataType.Type.TINYINT: "INT64",
692            exp.DataType.Type.VARBINARY: "BYTES",
693            exp.DataType.Type.VARCHAR: "STRING",
694            exp.DataType.Type.VARIANT: "ANY TYPE",
695        }
696
697        PROPERTIES_LOCATION = {
698            **generator.Generator.PROPERTIES_LOCATION,
699            exp.PartitionedByProperty: exp.Properties.Location.POST_SCHEMA,
700            exp.VolatileProperty: exp.Properties.Location.UNSUPPORTED,
701        }
702
703        # from: https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#reserved_keywords
704        RESERVED_KEYWORDS = {
705            *generator.Generator.RESERVED_KEYWORDS,
706            "all",
707            "and",
708            "any",
709            "array",
710            "as",
711            "asc",
712            "assert_rows_modified",
713            "at",
714            "between",
715            "by",
716            "case",
717            "cast",
718            "collate",
719            "contains",
720            "create",
721            "cross",
722            "cube",
723            "current",
724            "default",
725            "define",
726            "desc",
727            "distinct",
728            "else",
729            "end",
730            "enum",
731            "escape",
732            "except",
733            "exclude",
734            "exists",
735            "extract",
736            "false",
737            "fetch",
738            "following",
739            "for",
740            "from",
741            "full",
742            "group",
743            "grouping",
744            "groups",
745            "hash",
746            "having",
747            "if",
748            "ignore",
749            "in",
750            "inner",
751            "intersect",
752            "interval",
753            "into",
754            "is",
755            "join",
756            "lateral",
757            "left",
758            "like",
759            "limit",
760            "lookup",
761            "merge",
762            "natural",
763            "new",
764            "no",
765            "not",
766            "null",
767            "nulls",
768            "of",
769            "on",
770            "or",
771            "order",
772            "outer",
773            "over",
774            "partition",
775            "preceding",
776            "proto",
777            "qualify",
778            "range",
779            "recursive",
780            "respect",
781            "right",
782            "rollup",
783            "rows",
784            "select",
785            "set",
786            "some",
787            "struct",
788            "tablesample",
789            "then",
790            "to",
791            "treat",
792            "true",
793            "unbounded",
794            "union",
795            "unnest",
796            "using",
797            "when",
798            "where",
799            "window",
800            "with",
801            "within",
802        }
803
804        def table_parts(self, expression: exp.Table) -> str:
805            # Depending on the context, `x.y` may not resolve to the same data source as `x`.`y`, so
806            # we need to make sure the correct quoting is used in each case.
807            #
808            # For example, if there is a CTE x that clashes with a schema name, then the former will
809            # return the table y in that schema, whereas the latter will return the CTE's y column:
810            #
811            # - WITH x AS (SELECT [1, 2] AS y) SELECT * FROM x, `x.y`   -> cross join
812            # - WITH x AS (SELECT [1, 2] AS y) SELECT * FROM x, `x`.`y` -> implicit unnest
813            if expression.meta.get("quoted_table"):
814                table_parts = ".".join(p.name for p in expression.parts)
815                return self.sql(exp.Identifier(this=table_parts, quoted=True))
816
817            return super().table_parts(expression)
818
819        def timetostr_sql(self, expression: exp.TimeToStr) -> str:
820            this = expression.this if isinstance(expression.this, exp.TsOrDsToDate) else expression
821            return self.func("FORMAT_DATE", self.format_time(expression), this.this)
822
823        def eq_sql(self, expression: exp.EQ) -> str:
824            # Operands of = cannot be NULL in BigQuery
825            if isinstance(expression.left, exp.Null) or isinstance(expression.right, exp.Null):
826                if not isinstance(expression.parent, exp.Update):
827                    return "NULL"
828
829            return self.binary(expression, "=")
830
831        def attimezone_sql(self, expression: exp.AtTimeZone) -> str:
832            parent = expression.parent
833
834            # BigQuery allows CAST(.. AS {STRING|TIMESTAMP} [FORMAT <fmt> [AT TIME ZONE <tz>]]).
835            # Only the TIMESTAMP one should use the below conversion, when AT TIME ZONE is included.
836            if not isinstance(parent, exp.Cast) or not parent.to.is_type("text"):
837                return self.func(
838                    "TIMESTAMP", self.func("DATETIME", expression.this, expression.args.get("zone"))
839                )
840
841            return super().attimezone_sql(expression)
842
843        def trycast_sql(self, expression: exp.TryCast) -> str:
844            return self.cast_sql(expression, safe_prefix="SAFE_")
845
846        def array_sql(self, expression: exp.Array) -> str:
847            first_arg = seq_get(expression.expressions, 0)
848            if isinstance(first_arg, exp.Query):
849                return f"ARRAY{self.wrap(self.sql(first_arg))}"
850
851            return inline_array_sql(self, expression)
852
853        def bracket_sql(self, expression: exp.Bracket) -> str:
854            this = expression.this
855            expressions = expression.expressions
856
857            if len(expressions) == 1 and this and this.is_type(exp.DataType.Type.STRUCT):
858                arg = expressions[0]
859                if arg.type is None:
860                    from sqlglot.optimizer.annotate_types import annotate_types
861
862                    arg = annotate_types(arg)
863
864                if arg.type and arg.type.this in exp.DataType.TEXT_TYPES:
865                    # BQ doesn't support bracket syntax with string values for structs
866                    return f"{self.sql(this)}.{arg.name}"
867
868            expressions_sql = self.expressions(expression, flat=True)
869            offset = expression.args.get("offset")
870
871            if offset == 0:
872                expressions_sql = f"OFFSET({expressions_sql})"
873            elif offset == 1:
874                expressions_sql = f"ORDINAL({expressions_sql})"
875            elif offset is not None:
876                self.unsupported(f"Unsupported array offset: {offset}")
877
878            if expression.args.get("safe"):
879                expressions_sql = f"SAFE_{expressions_sql}"
880
881            return f"{self.sql(this)}[{expressions_sql}]"
882
883        def in_unnest_op(self, expression: exp.Unnest) -> str:
884            return self.sql(expression)
885
886        def except_op(self, expression: exp.Except) -> str:
887            if not expression.args.get("distinct"):
888                self.unsupported("EXCEPT without DISTINCT is not supported in BigQuery")
889            return f"EXCEPT{' DISTINCT' if expression.args.get('distinct') else ' ALL'}"
890
891        def intersect_op(self, expression: exp.Intersect) -> str:
892            if not expression.args.get("distinct"):
893                self.unsupported("INTERSECT without DISTINCT is not supported in BigQuery")
894            return f"INTERSECT{' DISTINCT' if expression.args.get('distinct') else ' ALL'}"
895
896        def with_properties(self, properties: exp.Properties) -> str:
897            return self.properties(properties, prefix=self.seg("OPTIONS"))
898
899        def version_sql(self, expression: exp.Version) -> str:
900            if expression.name == "TIMESTAMP":
901                expression.set("this", "SYSTEM_TIME")
902            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_UNION = 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 = '@'
TRANSFORMS = {<class 'sqlglot.expressions.JSONPathKey'>: <function <lambda>>, <class 'sqlglot.expressions.JSONPathRoot'>: <function <lambda>>, <class 'sqlglot.expressions.JSONPathSubscript'>: <function <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.DateAdd'>: <function date_add_interval_sql.<locals>.func>, <class 'sqlglot.expressions.DateFormatColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.DefaultColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.EncodeColumnConstraint'>: <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.RemoteWithConnectionModelProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ReturnsProperty'>: <function _returnsproperty_sql>, <class 'sqlglot.expressions.SampleProperty'>: <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.TemporaryProperty'>: <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.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.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 rename_func.<locals>.<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.Mod'>: <function rename_func.<locals>.<lambda>>, <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.SHA2'>: <function BigQuery.Generator.<lambda>>, <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.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.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.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.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.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.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.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 = {'fetch', 'asc', 'of', 'any', 'array', 'exclude', 'exists', 'nulls', 'order', 'rollup', 'qualify', 'except', 'group', 'set', 'into', 'end', 'having', 'lookup', 'and', 'at', 'over', 'inner', 'default', 'as', 'no', 'merge', 'interval', 'null', 'collate', 'all', 'new', 'tablesample', 'then', 'true', 'is', 'range', 'case', 'within', 'hash', 'respect', 'grouping', 'window', 'limit', 'distinct', 'outer', 'treat', 'when', 'unnest', 'some', 'where', 'union', 'else', 'create', 'cross', 'natural', 'cast', 'escape', 'select', 'by', 'false', 'partition', 'left', 'define', 'right', 'current', 'proto', 'assert_rows_modified', 'using', 'unbounded', 'on', 'desc', 'if', 'recursive', 'for', 'rows', 'extract', 'struct', 'or', 'from', 'with', 'cube', 'between', 'contains', 'in', 'to', 'preceding', 'like', 'groups', 'following', 'intersect', 'enum', 'full', 'lateral', 'not', 'ignore', 'join'}
def table_parts(self, expression: sqlglot.expressions.Table) -> str:
804        def table_parts(self, expression: exp.Table) -> str:
805            # Depending on the context, `x.y` may not resolve to the same data source as `x`.`y`, so
806            # we need to make sure the correct quoting is used in each case.
807            #
808            # For example, if there is a CTE x that clashes with a schema name, then the former will
809            # return the table y in that schema, whereas the latter will return the CTE's y column:
810            #
811            # - WITH x AS (SELECT [1, 2] AS y) SELECT * FROM x, `x.y`   -> cross join
812            # - WITH x AS (SELECT [1, 2] AS y) SELECT * FROM x, `x`.`y` -> implicit unnest
813            if expression.meta.get("quoted_table"):
814                table_parts = ".".join(p.name for p in expression.parts)
815                return self.sql(exp.Identifier(this=table_parts, quoted=True))
816
817            return super().table_parts(expression)
def timetostr_sql(self, expression: sqlglot.expressions.TimeToStr) -> str:
819        def timetostr_sql(self, expression: exp.TimeToStr) -> str:
820            this = expression.this if isinstance(expression.this, exp.TsOrDsToDate) else expression
821            return self.func("FORMAT_DATE", self.format_time(expression), this.this)
def eq_sql(self, expression: sqlglot.expressions.EQ) -> str:
823        def eq_sql(self, expression: exp.EQ) -> str:
824            # Operands of = cannot be NULL in BigQuery
825            if isinstance(expression.left, exp.Null) or isinstance(expression.right, exp.Null):
826                if not isinstance(expression.parent, exp.Update):
827                    return "NULL"
828
829            return self.binary(expression, "=")
def attimezone_sql(self, expression: sqlglot.expressions.AtTimeZone) -> str:
831        def attimezone_sql(self, expression: exp.AtTimeZone) -> str:
832            parent = expression.parent
833
834            # BigQuery allows CAST(.. AS {STRING|TIMESTAMP} [FORMAT <fmt> [AT TIME ZONE <tz>]]).
835            # Only the TIMESTAMP one should use the below conversion, when AT TIME ZONE is included.
836            if not isinstance(parent, exp.Cast) or not parent.to.is_type("text"):
837                return self.func(
838                    "TIMESTAMP", self.func("DATETIME", expression.this, expression.args.get("zone"))
839                )
840
841            return super().attimezone_sql(expression)
def trycast_sql(self, expression: sqlglot.expressions.TryCast) -> str:
843        def trycast_sql(self, expression: exp.TryCast) -> str:
844            return self.cast_sql(expression, safe_prefix="SAFE_")
def array_sql(self, expression: sqlglot.expressions.Array) -> str:
846        def array_sql(self, expression: exp.Array) -> str:
847            first_arg = seq_get(expression.expressions, 0)
848            if isinstance(first_arg, exp.Query):
849                return f"ARRAY{self.wrap(self.sql(first_arg))}"
850
851            return inline_array_sql(self, expression)
def bracket_sql(self, expression: sqlglot.expressions.Bracket) -> str:
853        def bracket_sql(self, expression: exp.Bracket) -> str:
854            this = expression.this
855            expressions = expression.expressions
856
857            if len(expressions) == 1 and this and this.is_type(exp.DataType.Type.STRUCT):
858                arg = expressions[0]
859                if arg.type is None:
860                    from sqlglot.optimizer.annotate_types import annotate_types
861
862                    arg = annotate_types(arg)
863
864                if arg.type and arg.type.this in exp.DataType.TEXT_TYPES:
865                    # BQ doesn't support bracket syntax with string values for structs
866                    return f"{self.sql(this)}.{arg.name}"
867
868            expressions_sql = self.expressions(expression, flat=True)
869            offset = expression.args.get("offset")
870
871            if offset == 0:
872                expressions_sql = f"OFFSET({expressions_sql})"
873            elif offset == 1:
874                expressions_sql = f"ORDINAL({expressions_sql})"
875            elif offset is not None:
876                self.unsupported(f"Unsupported array offset: {offset}")
877
878            if expression.args.get("safe"):
879                expressions_sql = f"SAFE_{expressions_sql}"
880
881            return f"{self.sql(this)}[{expressions_sql}]"
def in_unnest_op(self, expression: sqlglot.expressions.Unnest) -> str:
883        def in_unnest_op(self, expression: exp.Unnest) -> str:
884            return self.sql(expression)
def except_op(self, expression: sqlglot.expressions.Except) -> str:
886        def except_op(self, expression: exp.Except) -> str:
887            if not expression.args.get("distinct"):
888                self.unsupported("EXCEPT without DISTINCT is not supported in BigQuery")
889            return f"EXCEPT{' DISTINCT' if expression.args.get('distinct') else ' ALL'}"
def intersect_op(self, expression: sqlglot.expressions.Intersect) -> str:
891        def intersect_op(self, expression: exp.Intersect) -> str:
892            if not expression.args.get("distinct"):
893                self.unsupported("INTERSECT without DISTINCT is not supported in BigQuery")
894            return f"INTERSECT{' DISTINCT' if expression.args.get('distinct') else ' ALL'}"
def with_properties(self, properties: sqlglot.expressions.Properties) -> str:
896        def with_properties(self, properties: exp.Properties) -> str:
897            return self.properties(properties, prefix=self.seg("OPTIONS"))
def version_sql(self, expression: sqlglot.expressions.Version) -> str:
899        def version_sql(self, expression: exp.Version) -> str:
900            if expression.name == "TIMESTAMP":
901                expression.set("this", "SYSTEM_TIME")
902            return super().version_sql(expression)
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
COLUMN_JOIN_MARKS_SUPPORTED
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
STAR_MAPPING
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
inputoutputformat_sql
national_sql
partition_sql
properties_sql
root_properties
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
matchrecognize_sql
query_modifiers
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
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
renametable_sql
renamecolumn_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
mod_sql
mul_sql
neq_sql
nullsafeeq_sql
nullsafeneq_sql
slice_sql
sub_sql
log_sql
use_sql
binary
function_fallback_sql
func
format_args
text_width
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
arrayany_sql
generateseries_sql
struct_sql
partitionrange_sql
truncatetable_sql
convert_sql