Edit on GitHub

sqlglot.dialects.duckdb

  1from __future__ import annotations
  2
  3import typing as t
  4
  5from sqlglot import exp, generator, parser, tokens, transforms
  6from sqlglot.expressions import DATA_TYPE
  7from sqlglot.dialects.dialect import (
  8    Dialect,
  9    JSON_EXTRACT_TYPE,
 10    NormalizationStrategy,
 11    approx_count_distinct_sql,
 12    arg_max_or_min_no_count,
 13    arrow_json_extract_sql,
 14    binary_from_function,
 15    bool_xor_sql,
 16    build_default_decimal_type,
 17    date_trunc_to_time,
 18    datestrtodate_sql,
 19    no_datetime_sql,
 20    encode_decode_sql,
 21    build_formatted_time,
 22    inline_array_unless_query,
 23    no_comment_column_constraint_sql,
 24    no_safe_divide_sql,
 25    no_time_sql,
 26    no_timestamp_sql,
 27    pivot_column_names,
 28    rename_func,
 29    str_position_sql,
 30    str_to_time_sql,
 31    timestamptrunc_sql,
 32    timestrtotime_sql,
 33    unit_to_var,
 34    unit_to_str,
 35    sha256_sql,
 36    build_regexp_extract,
 37    explode_to_unnest_sql,
 38)
 39from sqlglot.generator import unsupported_args
 40from sqlglot.helper import seq_get
 41from sqlglot.tokens import TokenType
 42from sqlglot.parser import binary_range_parser
 43
 44DATETIME_DELTA = t.Union[
 45    exp.DateAdd, exp.TimeAdd, exp.DatetimeAdd, exp.TsOrDsAdd, exp.DateSub, exp.DatetimeSub
 46]
 47
 48WINDOW_FUNCS_WITH_IGNORE_NULLS = (
 49    exp.FirstValue,
 50    exp.LastValue,
 51    exp.Lag,
 52    exp.Lead,
 53    exp.NthValue,
 54)
 55
 56
 57def _date_delta_sql(self: DuckDB.Generator, expression: DATETIME_DELTA) -> str:
 58    this = expression.this
 59    unit = unit_to_var(expression)
 60    op = (
 61        "+"
 62        if isinstance(expression, (exp.DateAdd, exp.TimeAdd, exp.DatetimeAdd, exp.TsOrDsAdd))
 63        else "-"
 64    )
 65
 66    to_type: t.Optional[DATA_TYPE] = None
 67    if isinstance(expression, exp.TsOrDsAdd):
 68        to_type = expression.return_type
 69    elif this.is_string:
 70        # Cast string literals (i.e function parameters) to the appropriate type for +/- interval to work
 71        to_type = (
 72            exp.DataType.Type.DATETIME
 73            if isinstance(expression, (exp.DatetimeAdd, exp.DatetimeSub))
 74            else exp.DataType.Type.DATE
 75        )
 76
 77    this = exp.cast(this, to_type) if to_type else this
 78
 79    return f"{self.sql(this)} {op} {self.sql(exp.Interval(this=expression.expression, unit=unit))}"
 80
 81
 82# BigQuery -> DuckDB conversion for the DATE function
 83def _date_sql(self: DuckDB.Generator, expression: exp.Date) -> str:
 84    result = f"CAST({self.sql(expression, 'this')} AS DATE)"
 85    zone = self.sql(expression, "zone")
 86
 87    if zone:
 88        date_str = self.func("STRFTIME", result, "'%d/%m/%Y'")
 89        date_str = f"{date_str} || ' ' || {zone}"
 90
 91        # This will create a TIMESTAMP with time zone information
 92        result = self.func("STRPTIME", date_str, "'%d/%m/%Y %Z'")
 93
 94    return result
 95
 96
 97# BigQuery -> DuckDB conversion for the TIME_DIFF function
 98def _timediff_sql(self: DuckDB.Generator, expression: exp.TimeDiff) -> str:
 99    this = exp.cast(expression.this, exp.DataType.Type.TIME)
100    expr = exp.cast(expression.expression, exp.DataType.Type.TIME)
101
102    # Although the 2 dialects share similar signatures, BQ seems to inverse
103    # the sign of the result so the start/end time operands are flipped
104    return self.func("DATE_DIFF", unit_to_str(expression), expr, this)
105
106
107@unsupported_args(("expression", "DuckDB's ARRAY_SORT does not support a comparator."))
108def _array_sort_sql(self: DuckDB.Generator, expression: exp.ArraySort) -> str:
109    return self.func("ARRAY_SORT", expression.this)
110
111
112def _sort_array_sql(self: DuckDB.Generator, expression: exp.SortArray) -> str:
113    name = "ARRAY_REVERSE_SORT" if expression.args.get("asc") == exp.false() else "ARRAY_SORT"
114    return self.func(name, expression.this)
115
116
117def _build_sort_array_desc(args: t.List) -> exp.Expression:
118    return exp.SortArray(this=seq_get(args, 0), asc=exp.false())
119
120
121def _build_date_diff(args: t.List) -> exp.Expression:
122    return exp.DateDiff(this=seq_get(args, 2), expression=seq_get(args, 1), unit=seq_get(args, 0))
123
124
125def _build_generate_series(end_exclusive: bool = False) -> t.Callable[[t.List], exp.GenerateSeries]:
126    def _builder(args: t.List) -> exp.GenerateSeries:
127        # Check https://duckdb.org/docs/sql/functions/nested.html#range-functions
128        if len(args) == 1:
129            # DuckDB uses 0 as a default for the series' start when it's omitted
130            args.insert(0, exp.Literal.number("0"))
131
132        gen_series = exp.GenerateSeries.from_arg_list(args)
133        gen_series.set("is_end_exclusive", end_exclusive)
134
135        return gen_series
136
137    return _builder
138
139
140def _build_make_timestamp(args: t.List) -> exp.Expression:
141    if len(args) == 1:
142        return exp.UnixToTime(this=seq_get(args, 0), scale=exp.UnixToTime.MICROS)
143
144    return exp.TimestampFromParts(
145        year=seq_get(args, 0),
146        month=seq_get(args, 1),
147        day=seq_get(args, 2),
148        hour=seq_get(args, 3),
149        min=seq_get(args, 4),
150        sec=seq_get(args, 5),
151    )
152
153
154def _struct_sql(self: DuckDB.Generator, expression: exp.Struct) -> str:
155    args: t.List[str] = []
156
157    # BigQuery allows inline construction such as "STRUCT<a STRING, b INTEGER>('str', 1)" which is
158    # canonicalized to "ROW('str', 1) AS STRUCT(a TEXT, b INT)" in DuckDB
159    # The transformation to ROW will take place if:
160    #  1. The STRUCT itself does not have proper fields (key := value) as a "proper" STRUCT would
161    #  2. A cast to STRUCT / ARRAY of STRUCTs is found
162    ancestor_cast = expression.find_ancestor(exp.Cast)
163    is_bq_inline_struct = (
164        (expression.find(exp.PropertyEQ) is None)
165        and ancestor_cast
166        and any(
167            casted_type.is_type(exp.DataType.Type.STRUCT)
168            for casted_type in ancestor_cast.find_all(exp.DataType)
169        )
170    )
171
172    for i, expr in enumerate(expression.expressions):
173        is_property_eq = isinstance(expr, exp.PropertyEQ)
174        value = expr.expression if is_property_eq else expr
175
176        if is_bq_inline_struct:
177            args.append(self.sql(value))
178        else:
179            key = expr.name if is_property_eq else f"_{i}"
180            args.append(f"{self.sql(exp.Literal.string(key))}: {self.sql(value)}")
181
182    csv_args = ", ".join(args)
183
184    return f"ROW({csv_args})" if is_bq_inline_struct else f"{{{csv_args}}}"
185
186
187def _datatype_sql(self: DuckDB.Generator, expression: exp.DataType) -> str:
188    if expression.is_type("array"):
189        return f"{self.expressions(expression, flat=True)}[{self.expressions(expression, key='values', flat=True)}]"
190
191    # Modifiers are not supported for TIME, [TIME | TIMESTAMP] WITH TIME ZONE
192    if expression.is_type(
193        exp.DataType.Type.TIME, exp.DataType.Type.TIMETZ, exp.DataType.Type.TIMESTAMPTZ
194    ):
195        return expression.this.value
196
197    return self.datatype_sql(expression)
198
199
200def _json_format_sql(self: DuckDB.Generator, expression: exp.JSONFormat) -> str:
201    sql = self.func("TO_JSON", expression.this, expression.args.get("options"))
202    return f"CAST({sql} AS TEXT)"
203
204
205def _unix_to_time_sql(self: DuckDB.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("TO_TIMESTAMP", timestamp)
211    if scale == exp.UnixToTime.MILLIS:
212        return self.func("EPOCH_MS", timestamp)
213    if scale == exp.UnixToTime.MICROS:
214        return self.func("MAKE_TIMESTAMP", timestamp)
215
216    return self.func("TO_TIMESTAMP", exp.Div(this=timestamp, expression=exp.func("POW", 10, scale)))
217
218
219WRAPPED_JSON_EXTRACT_EXPRESSIONS = (exp.Binary, exp.Bracket, exp.In)
220
221
222def _arrow_json_extract_sql(self: DuckDB.Generator, expression: JSON_EXTRACT_TYPE) -> str:
223    arrow_sql = arrow_json_extract_sql(self, expression)
224    if not expression.same_parent and isinstance(
225        expression.parent, WRAPPED_JSON_EXTRACT_EXPRESSIONS
226    ):
227        arrow_sql = self.wrap(arrow_sql)
228    return arrow_sql
229
230
231def _implicit_datetime_cast(
232    arg: t.Optional[exp.Expression], type: exp.DataType.Type = exp.DataType.Type.DATE
233) -> t.Optional[exp.Expression]:
234    return exp.cast(arg, type) if isinstance(arg, exp.Literal) else arg
235
236
237def _date_diff_sql(self: DuckDB.Generator, expression: exp.DateDiff) -> str:
238    this = _implicit_datetime_cast(expression.this)
239    expr = _implicit_datetime_cast(expression.expression)
240
241    return self.func("DATE_DIFF", unit_to_str(expression), expr, this)
242
243
244def _generate_datetime_array_sql(
245    self: DuckDB.Generator, expression: t.Union[exp.GenerateDateArray, exp.GenerateTimestampArray]
246) -> str:
247    is_generate_date_array = isinstance(expression, exp.GenerateDateArray)
248
249    type = exp.DataType.Type.DATE if is_generate_date_array else exp.DataType.Type.TIMESTAMP
250    start = _implicit_datetime_cast(expression.args.get("start"), type=type)
251    end = _implicit_datetime_cast(expression.args.get("end"), type=type)
252
253    # BQ's GENERATE_DATE_ARRAY & GENERATE_TIMESTAMP_ARRAY are transformed to DuckDB'S GENERATE_SERIES
254    gen_series: t.Union[exp.GenerateSeries, exp.Cast] = exp.GenerateSeries(
255        start=start, end=end, step=expression.args.get("step")
256    )
257
258    if is_generate_date_array:
259        # The GENERATE_SERIES result type is TIMESTAMP array, so to match BQ's semantics for
260        # GENERATE_DATE_ARRAY we must cast it back to DATE array
261        gen_series = exp.cast(gen_series, exp.DataType.build("ARRAY<DATE>"))
262
263    return self.sql(gen_series)
264
265
266def _json_extract_value_array_sql(
267    self: DuckDB.Generator, expression: exp.JSONValueArray | exp.JSONExtractArray
268) -> str:
269    json_extract = exp.JSONExtract(this=expression.this, expression=expression.expression)
270    data_type = "ARRAY<STRING>" if isinstance(expression, exp.JSONValueArray) else "ARRAY<JSON>"
271    return self.sql(exp.cast(json_extract, to=exp.DataType.build(data_type)))
272
273
274class DuckDB(Dialect):
275    NULL_ORDERING = "nulls_are_last"
276    SUPPORTS_USER_DEFINED_TYPES = False
277    SAFE_DIVISION = True
278    INDEX_OFFSET = 1
279    CONCAT_COALESCE = True
280    SUPPORTS_ORDER_BY_ALL = True
281    SUPPORTS_FIXED_SIZE_ARRAYS = True
282    STRICT_JSON_PATH_SYNTAX = False
283
284    # https://duckdb.org/docs/sql/introduction.html#creating-a-new-table
285    NORMALIZATION_STRATEGY = NormalizationStrategy.CASE_INSENSITIVE
286
287    def to_json_path(self, path: t.Optional[exp.Expression]) -> t.Optional[exp.Expression]:
288        if isinstance(path, exp.Literal):
289            # DuckDB also supports the JSON pointer syntax, where every path starts with a `/`.
290            # Additionally, it allows accessing the back of lists using the `[#-i]` syntax.
291            # This check ensures we'll avoid trying to parse these as JSON paths, which can
292            # either result in a noisy warning or in an invalid representation of the path.
293            path_text = path.name
294            if path_text.startswith("/") or "[#" in path_text:
295                return path
296
297        return super().to_json_path(path)
298
299    class Tokenizer(tokens.Tokenizer):
300        HEREDOC_STRINGS = ["$"]
301
302        HEREDOC_TAG_IS_IDENTIFIER = True
303        HEREDOC_STRING_ALTERNATIVE = TokenType.PARAMETER
304
305        KEYWORDS = {
306            **tokens.Tokenizer.KEYWORDS,
307            "//": TokenType.DIV,
308            "**": TokenType.DSTAR,
309            "^@": TokenType.CARET_AT,
310            "@>": TokenType.AT_GT,
311            "<@": TokenType.LT_AT,
312            "ATTACH": TokenType.COMMAND,
313            "BINARY": TokenType.VARBINARY,
314            "BITSTRING": TokenType.BIT,
315            "BPCHAR": TokenType.TEXT,
316            "CHAR": TokenType.TEXT,
317            "CHARACTER VARYING": TokenType.TEXT,
318            "EXCLUDE": TokenType.EXCEPT,
319            "LOGICAL": TokenType.BOOLEAN,
320            "ONLY": TokenType.ONLY,
321            "PIVOT_WIDER": TokenType.PIVOT,
322            "POSITIONAL": TokenType.POSITIONAL,
323            "SIGNED": TokenType.INT,
324            "STRING": TokenType.TEXT,
325            "SUMMARIZE": TokenType.SUMMARIZE,
326            "TIMESTAMP_S": TokenType.TIMESTAMP_S,
327            "TIMESTAMP_MS": TokenType.TIMESTAMP_MS,
328            "TIMESTAMP_NS": TokenType.TIMESTAMP_NS,
329            "TIMESTAMP_US": TokenType.TIMESTAMP,
330            "UBIGINT": TokenType.UBIGINT,
331            "UINTEGER": TokenType.UINT,
332            "USMALLINT": TokenType.USMALLINT,
333            "UTINYINT": TokenType.UTINYINT,
334            "VARCHAR": TokenType.TEXT,
335        }
336        KEYWORDS.pop("/*+")
337
338        SINGLE_TOKENS = {
339            **tokens.Tokenizer.SINGLE_TOKENS,
340            "$": TokenType.PARAMETER,
341        }
342
343    class Parser(parser.Parser):
344        BITWISE = {
345            **parser.Parser.BITWISE,
346            TokenType.TILDA: exp.RegexpLike,
347        }
348        BITWISE.pop(TokenType.CARET)
349
350        RANGE_PARSERS = {
351            **parser.Parser.RANGE_PARSERS,
352            TokenType.DAMP: binary_range_parser(exp.ArrayOverlaps),
353            TokenType.CARET_AT: binary_range_parser(exp.StartsWith),
354        }
355
356        EXPONENT = {
357            **parser.Parser.EXPONENT,
358            TokenType.CARET: exp.Pow,
359            TokenType.DSTAR: exp.Pow,
360        }
361
362        FUNCTIONS_WITH_ALIASED_ARGS = {*parser.Parser.FUNCTIONS_WITH_ALIASED_ARGS, "STRUCT_PACK"}
363
364        FUNCTIONS = {
365            **parser.Parser.FUNCTIONS,
366            "ARRAY_REVERSE_SORT": _build_sort_array_desc,
367            "ARRAY_SORT": exp.SortArray.from_arg_list,
368            "DATEDIFF": _build_date_diff,
369            "DATE_DIFF": _build_date_diff,
370            "DATE_TRUNC": date_trunc_to_time,
371            "DATETRUNC": date_trunc_to_time,
372            "DECODE": lambda args: exp.Decode(
373                this=seq_get(args, 0), charset=exp.Literal.string("utf-8")
374            ),
375            "ENCODE": lambda args: exp.Encode(
376                this=seq_get(args, 0), charset=exp.Literal.string("utf-8")
377            ),
378            "EPOCH": exp.TimeToUnix.from_arg_list,
379            "EPOCH_MS": lambda args: exp.UnixToTime(
380                this=seq_get(args, 0), scale=exp.UnixToTime.MILLIS
381            ),
382            "JSON": exp.ParseJSON.from_arg_list,
383            "JSON_EXTRACT_PATH": parser.build_extract_json_with_path(exp.JSONExtract),
384            "JSON_EXTRACT_STRING": parser.build_extract_json_with_path(exp.JSONExtractScalar),
385            "LIST_HAS": exp.ArrayContains.from_arg_list,
386            "LIST_REVERSE_SORT": _build_sort_array_desc,
387            "LIST_SORT": exp.SortArray.from_arg_list,
388            "LIST_VALUE": lambda args: exp.Array(expressions=args),
389            "MAKE_TIME": exp.TimeFromParts.from_arg_list,
390            "MAKE_TIMESTAMP": _build_make_timestamp,
391            "QUANTILE_CONT": exp.PercentileCont.from_arg_list,
392            "QUANTILE_DISC": exp.PercentileDisc.from_arg_list,
393            "REGEXP_EXTRACT": build_regexp_extract(exp.RegexpExtract),
394            "REGEXP_EXTRACT_ALL": build_regexp_extract(exp.RegexpExtractAll),
395            "REGEXP_MATCHES": exp.RegexpLike.from_arg_list,
396            "REGEXP_REPLACE": lambda args: exp.RegexpReplace(
397                this=seq_get(args, 0),
398                expression=seq_get(args, 1),
399                replacement=seq_get(args, 2),
400                modifiers=seq_get(args, 3),
401            ),
402            "STRFTIME": build_formatted_time(exp.TimeToStr, "duckdb"),
403            "STRING_SPLIT": exp.Split.from_arg_list,
404            "STRING_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
405            "STRING_TO_ARRAY": exp.Split.from_arg_list,
406            "STRPTIME": build_formatted_time(exp.StrToTime, "duckdb"),
407            "STRUCT_PACK": exp.Struct.from_arg_list,
408            "STR_SPLIT": exp.Split.from_arg_list,
409            "STR_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
410            "TO_TIMESTAMP": exp.UnixToTime.from_arg_list,
411            "UNNEST": exp.Explode.from_arg_list,
412            "XOR": binary_from_function(exp.BitwiseXor),
413            "GENERATE_SERIES": _build_generate_series(),
414            "RANGE": _build_generate_series(end_exclusive=True),
415            "EDITDIST3": exp.Levenshtein.from_arg_list,
416        }
417
418        FUNCTIONS.pop("DATE_SUB")
419        FUNCTIONS.pop("GLOB")
420
421        FUNCTION_PARSERS = parser.Parser.FUNCTION_PARSERS.copy()
422        FUNCTION_PARSERS.pop("DECODE")
423
424        NO_PAREN_FUNCTION_PARSERS = {
425            **parser.Parser.NO_PAREN_FUNCTION_PARSERS,
426            "MAP": lambda self: self._parse_map(),
427        }
428
429        TABLE_ALIAS_TOKENS = parser.Parser.TABLE_ALIAS_TOKENS - {
430            TokenType.SEMI,
431            TokenType.ANTI,
432        }
433
434        PLACEHOLDER_PARSERS = {
435            **parser.Parser.PLACEHOLDER_PARSERS,
436            TokenType.PARAMETER: lambda self: (
437                self.expression(exp.Placeholder, this=self._prev.text)
438                if self._match(TokenType.NUMBER) or self._match_set(self.ID_VAR_TOKENS)
439                else None
440            ),
441        }
442
443        TYPE_CONVERTERS = {
444            # https://duckdb.org/docs/sql/data_types/numeric
445            exp.DataType.Type.DECIMAL: build_default_decimal_type(precision=18, scale=3),
446            # https://duckdb.org/docs/sql/data_types/text
447            exp.DataType.Type.TEXT: lambda dtype: exp.DataType.build("TEXT"),
448        }
449
450        def _parse_table_sample(self, as_modifier: bool = False) -> t.Optional[exp.TableSample]:
451            # https://duckdb.org/docs/sql/samples.html
452            sample = super()._parse_table_sample(as_modifier=as_modifier)
453            if sample and not sample.args.get("method"):
454                if sample.args.get("size"):
455                    sample.set("method", exp.var("RESERVOIR"))
456                else:
457                    sample.set("method", exp.var("SYSTEM"))
458
459            return sample
460
461        def _parse_bracket(
462            self, this: t.Optional[exp.Expression] = None
463        ) -> t.Optional[exp.Expression]:
464            bracket = super()._parse_bracket(this)
465            if isinstance(bracket, exp.Bracket):
466                bracket.set("returns_list_for_maps", True)
467
468            return bracket
469
470        def _parse_map(self) -> exp.ToMap | exp.Map:
471            if self._match(TokenType.L_BRACE, advance=False):
472                return self.expression(exp.ToMap, this=self._parse_bracket())
473
474            args = self._parse_wrapped_csv(self._parse_assignment)
475            return self.expression(exp.Map, keys=seq_get(args, 0), values=seq_get(args, 1))
476
477        def _parse_struct_types(self, type_required: bool = False) -> t.Optional[exp.Expression]:
478            return self._parse_field_def()
479
480        def _pivot_column_names(self, aggregations: t.List[exp.Expression]) -> t.List[str]:
481            if len(aggregations) == 1:
482                return super()._pivot_column_names(aggregations)
483            return pivot_column_names(aggregations, dialect="duckdb")
484
485    class Generator(generator.Generator):
486        PARAMETER_TOKEN = "$"
487        NAMED_PLACEHOLDER_TOKEN = "$"
488        JOIN_HINTS = False
489        TABLE_HINTS = False
490        QUERY_HINTS = False
491        LIMIT_FETCH = "LIMIT"
492        STRUCT_DELIMITER = ("(", ")")
493        RENAME_TABLE_WITH_DB = False
494        NVL2_SUPPORTED = False
495        SEMI_ANTI_JOIN_WITH_SIDE = False
496        TABLESAMPLE_KEYWORDS = "USING SAMPLE"
497        TABLESAMPLE_SEED_KEYWORD = "REPEATABLE"
498        LAST_DAY_SUPPORTS_DATE_PART = False
499        JSON_KEY_VALUE_PAIR_SEP = ","
500        IGNORE_NULLS_IN_FUNC = True
501        JSON_PATH_BRACKETED_KEY_SUPPORTED = False
502        SUPPORTS_CREATE_TABLE_LIKE = False
503        MULTI_ARG_DISTINCT = False
504        CAN_IMPLEMENT_ARRAY_ANY = True
505        SUPPORTS_TO_NUMBER = False
506        COPY_HAS_INTO_KEYWORD = False
507        STAR_EXCEPT = "EXCLUDE"
508        PAD_FILL_PATTERN_IS_REQUIRED = True
509        ARRAY_CONCAT_IS_VAR_LEN = False
510        ARRAY_SIZE_DIM_REQUIRED = False
511
512        TRANSFORMS = {
513            **generator.Generator.TRANSFORMS,
514            exp.ApproxDistinct: approx_count_distinct_sql,
515            exp.Array: inline_array_unless_query,
516            exp.ArrayFilter: rename_func("LIST_FILTER"),
517            exp.ArgMax: arg_max_or_min_no_count("ARG_MAX"),
518            exp.ArgMin: arg_max_or_min_no_count("ARG_MIN"),
519            exp.ArraySort: _array_sort_sql,
520            exp.ArraySum: rename_func("LIST_SUM"),
521            exp.BitwiseXor: rename_func("XOR"),
522            exp.CommentColumnConstraint: no_comment_column_constraint_sql,
523            exp.CurrentDate: lambda *_: "CURRENT_DATE",
524            exp.CurrentTime: lambda *_: "CURRENT_TIME",
525            exp.CurrentTimestamp: lambda *_: "CURRENT_TIMESTAMP",
526            exp.DayOfMonth: rename_func("DAYOFMONTH"),
527            exp.DayOfWeek: rename_func("DAYOFWEEK"),
528            exp.DayOfWeekIso: rename_func("ISODOW"),
529            exp.DayOfYear: rename_func("DAYOFYEAR"),
530            exp.DataType: _datatype_sql,
531            exp.Date: _date_sql,
532            exp.DateAdd: _date_delta_sql,
533            exp.DateFromParts: rename_func("MAKE_DATE"),
534            exp.DateSub: _date_delta_sql,
535            exp.DateDiff: _date_diff_sql,
536            exp.DateStrToDate: datestrtodate_sql,
537            exp.Datetime: no_datetime_sql,
538            exp.DatetimeSub: _date_delta_sql,
539            exp.DatetimeAdd: _date_delta_sql,
540            exp.DateToDi: lambda self,
541            e: f"CAST(STRFTIME({self.sql(e, 'this')}, {DuckDB.DATEINT_FORMAT}) AS INT)",
542            exp.Decode: lambda self, e: encode_decode_sql(self, e, "DECODE", replace=False),
543            exp.DiToDate: lambda self,
544            e: f"CAST(STRPTIME(CAST({self.sql(e, 'this')} AS TEXT), {DuckDB.DATEINT_FORMAT}) AS DATE)",
545            exp.Encode: lambda self, e: encode_decode_sql(self, e, "ENCODE", replace=False),
546            exp.GenerateDateArray: _generate_datetime_array_sql,
547            exp.GenerateTimestampArray: _generate_datetime_array_sql,
548            exp.Explode: rename_func("UNNEST"),
549            exp.IntDiv: lambda self, e: self.binary(e, "//"),
550            exp.IsInf: rename_func("ISINF"),
551            exp.IsNan: rename_func("ISNAN"),
552            exp.JSONBExists: rename_func("JSON_EXISTS"),
553            exp.JSONExtract: _arrow_json_extract_sql,
554            exp.JSONExtractArray: _json_extract_value_array_sql,
555            exp.JSONExtractScalar: _arrow_json_extract_sql,
556            exp.JSONFormat: _json_format_sql,
557            exp.JSONValueArray: _json_extract_value_array_sql,
558            exp.Lateral: explode_to_unnest_sql,
559            exp.LogicalOr: rename_func("BOOL_OR"),
560            exp.LogicalAnd: rename_func("BOOL_AND"),
561            exp.MD5Digest: lambda self, e: self.func("UNHEX", self.func("MD5", e.this)),
562            exp.MonthsBetween: lambda self, e: self.func(
563                "DATEDIFF",
564                "'month'",
565                exp.cast(e.expression, exp.DataType.Type.TIMESTAMP, copy=True),
566                exp.cast(e.this, exp.DataType.Type.TIMESTAMP, copy=True),
567            ),
568            exp.PercentileCont: rename_func("QUANTILE_CONT"),
569            exp.PercentileDisc: rename_func("QUANTILE_DISC"),
570            # DuckDB doesn't allow qualified columns inside of PIVOT expressions.
571            # See: https://github.com/duckdb/duckdb/blob/671faf92411182f81dce42ac43de8bfb05d9909e/src/planner/binder/tableref/bind_pivot.cpp#L61-L62
572            exp.Pivot: transforms.preprocess([transforms.unqualify_columns]),
573            exp.RegexpReplace: lambda self, e: self.func(
574                "REGEXP_REPLACE",
575                e.this,
576                e.expression,
577                e.args.get("replacement"),
578                e.args.get("modifiers"),
579            ),
580            exp.RegexpLike: rename_func("REGEXP_MATCHES"),
581            exp.RegexpSplit: rename_func("STR_SPLIT_REGEX"),
582            exp.Return: lambda self, e: self.sql(e, "this"),
583            exp.ReturnsProperty: lambda self, e: "TABLE" if isinstance(e.this, exp.Schema) else "",
584            exp.Rand: rename_func("RANDOM"),
585            exp.SafeDivide: no_safe_divide_sql,
586            exp.SHA: rename_func("SHA1"),
587            exp.SHA2: sha256_sql,
588            exp.Split: rename_func("STR_SPLIT"),
589            exp.SortArray: _sort_array_sql,
590            exp.StrPosition: str_position_sql,
591            exp.StrToUnix: lambda self, e: self.func(
592                "EPOCH", self.func("STRPTIME", e.this, self.format_time(e))
593            ),
594            exp.Struct: _struct_sql,
595            exp.Transform: rename_func("LIST_TRANSFORM"),
596            exp.TimeAdd: _date_delta_sql,
597            exp.Time: no_time_sql,
598            exp.TimeDiff: _timediff_sql,
599            exp.Timestamp: no_timestamp_sql,
600            exp.TimestampDiff: lambda self, e: self.func(
601                "DATE_DIFF", exp.Literal.string(e.unit), e.expression, e.this
602            ),
603            exp.TimestampTrunc: timestamptrunc_sql(),
604            exp.TimeStrToDate: lambda self, e: self.sql(exp.cast(e.this, exp.DataType.Type.DATE)),
605            exp.TimeStrToTime: timestrtotime_sql,
606            exp.TimeStrToUnix: lambda self, e: self.func(
607                "EPOCH", exp.cast(e.this, exp.DataType.Type.TIMESTAMP)
608            ),
609            exp.TimeToStr: lambda self, e: self.func("STRFTIME", e.this, self.format_time(e)),
610            exp.TimeToUnix: rename_func("EPOCH"),
611            exp.TsOrDiToDi: lambda self,
612            e: f"CAST(SUBSTR(REPLACE(CAST({self.sql(e, 'this')} AS TEXT), '-', ''), 1, 8) AS INT)",
613            exp.TsOrDsAdd: _date_delta_sql,
614            exp.TsOrDsDiff: lambda self, e: self.func(
615                "DATE_DIFF",
616                f"'{e.args.get('unit') or 'DAY'}'",
617                exp.cast(e.expression, exp.DataType.Type.TIMESTAMP),
618                exp.cast(e.this, exp.DataType.Type.TIMESTAMP),
619            ),
620            exp.UnixToStr: lambda self, e: self.func(
621                "STRFTIME", self.func("TO_TIMESTAMP", e.this), self.format_time(e)
622            ),
623            exp.DatetimeTrunc: lambda self, e: self.func(
624                "DATE_TRUNC", unit_to_str(e), exp.cast(e.this, exp.DataType.Type.DATETIME)
625            ),
626            exp.UnixToTime: _unix_to_time_sql,
627            exp.UnixToTimeStr: lambda self, e: f"CAST(TO_TIMESTAMP({self.sql(e, 'this')}) AS TEXT)",
628            exp.VariancePop: rename_func("VAR_POP"),
629            exp.WeekOfYear: rename_func("WEEKOFYEAR"),
630            exp.Xor: bool_xor_sql,
631            exp.Levenshtein: unsupported_args("ins_cost", "del_cost", "sub_cost", "max_dist")(
632                rename_func("LEVENSHTEIN")
633            ),
634        }
635
636        SUPPORTED_JSON_PATH_PARTS = {
637            exp.JSONPathKey,
638            exp.JSONPathRoot,
639            exp.JSONPathSubscript,
640            exp.JSONPathWildcard,
641        }
642
643        TYPE_MAPPING = {
644            **generator.Generator.TYPE_MAPPING,
645            exp.DataType.Type.BINARY: "BLOB",
646            exp.DataType.Type.BPCHAR: "TEXT",
647            exp.DataType.Type.CHAR: "TEXT",
648            exp.DataType.Type.FLOAT: "REAL",
649            exp.DataType.Type.NCHAR: "TEXT",
650            exp.DataType.Type.NVARCHAR: "TEXT",
651            exp.DataType.Type.UINT: "UINTEGER",
652            exp.DataType.Type.VARBINARY: "BLOB",
653            exp.DataType.Type.ROWVERSION: "BLOB",
654            exp.DataType.Type.VARCHAR: "TEXT",
655            exp.DataType.Type.TIMESTAMPNTZ: "TIMESTAMP",
656            exp.DataType.Type.TIMESTAMP_S: "TIMESTAMP_S",
657            exp.DataType.Type.TIMESTAMP_MS: "TIMESTAMP_MS",
658            exp.DataType.Type.TIMESTAMP_NS: "TIMESTAMP_NS",
659        }
660
661        # https://github.com/duckdb/duckdb/blob/ff7f24fd8e3128d94371827523dae85ebaf58713/third_party/libpg_query/grammar/keywords/reserved_keywords.list#L1-L77
662        RESERVED_KEYWORDS = {
663            "array",
664            "analyse",
665            "union",
666            "all",
667            "when",
668            "in_p",
669            "default",
670            "create_p",
671            "window",
672            "asymmetric",
673            "to",
674            "else",
675            "localtime",
676            "from",
677            "end_p",
678            "select",
679            "current_date",
680            "foreign",
681            "with",
682            "grant",
683            "session_user",
684            "or",
685            "except",
686            "references",
687            "fetch",
688            "limit",
689            "group_p",
690            "leading",
691            "into",
692            "collate",
693            "offset",
694            "do",
695            "then",
696            "localtimestamp",
697            "check_p",
698            "lateral_p",
699            "current_role",
700            "where",
701            "asc_p",
702            "placing",
703            "desc_p",
704            "user",
705            "unique",
706            "initially",
707            "column",
708            "both",
709            "some",
710            "as",
711            "any",
712            "only",
713            "deferrable",
714            "null_p",
715            "current_time",
716            "true_p",
717            "table",
718            "case",
719            "trailing",
720            "variadic",
721            "for",
722            "on",
723            "distinct",
724            "false_p",
725            "not",
726            "constraint",
727            "current_timestamp",
728            "returning",
729            "primary",
730            "intersect",
731            "having",
732            "analyze",
733            "current_user",
734            "and",
735            "cast",
736            "symmetric",
737            "using",
738            "order",
739            "current_catalog",
740        }
741
742        UNWRAPPED_INTERVAL_VALUES = (exp.Literal, exp.Paren)
743
744        # DuckDB doesn't generally support CREATE TABLE .. properties
745        # https://duckdb.org/docs/sql/statements/create_table.html
746        PROPERTIES_LOCATION = {
747            prop: exp.Properties.Location.UNSUPPORTED
748            for prop in generator.Generator.PROPERTIES_LOCATION
749        }
750
751        # There are a few exceptions (e.g. temporary tables) which are supported or
752        # can be transpiled to DuckDB, so we explicitly override them accordingly
753        PROPERTIES_LOCATION[exp.LikeProperty] = exp.Properties.Location.POST_SCHEMA
754        PROPERTIES_LOCATION[exp.TemporaryProperty] = exp.Properties.Location.POST_CREATE
755        PROPERTIES_LOCATION[exp.ReturnsProperty] = exp.Properties.Location.POST_ALIAS
756
757        def fromiso8601timestamp_sql(self, expression: exp.FromISO8601Timestamp) -> str:
758            return self.sql(exp.cast(expression.this, exp.DataType.Type.TIMESTAMPTZ))
759
760        def strtotime_sql(self, expression: exp.StrToTime) -> str:
761            if expression.args.get("safe"):
762                formatted_time = self.format_time(expression)
763                return f"CAST({self.func('TRY_STRPTIME', expression.this, formatted_time)} AS TIMESTAMP)"
764            return str_to_time_sql(self, expression)
765
766        def strtodate_sql(self, expression: exp.StrToDate) -> str:
767            if expression.args.get("safe"):
768                formatted_time = self.format_time(expression)
769                return f"CAST({self.func('TRY_STRPTIME', expression.this, formatted_time)} AS DATE)"
770            return f"CAST({str_to_time_sql(self, expression)} AS DATE)"
771
772        def parsejson_sql(self, expression: exp.ParseJSON) -> str:
773            arg = expression.this
774            if expression.args.get("safe"):
775                return self.sql(exp.case().when(exp.func("json_valid", arg), arg).else_(exp.null()))
776            return self.func("JSON", arg)
777
778        def timefromparts_sql(self, expression: exp.TimeFromParts) -> str:
779            nano = expression.args.get("nano")
780            if nano is not None:
781                expression.set(
782                    "sec", expression.args["sec"] + nano.pop() / exp.Literal.number(1000000000.0)
783                )
784
785            return rename_func("MAKE_TIME")(self, expression)
786
787        def timestampfromparts_sql(self, expression: exp.TimestampFromParts) -> str:
788            sec = expression.args["sec"]
789
790            milli = expression.args.get("milli")
791            if milli is not None:
792                sec += milli.pop() / exp.Literal.number(1000.0)
793
794            nano = expression.args.get("nano")
795            if nano is not None:
796                sec += nano.pop() / exp.Literal.number(1000000000.0)
797
798            if milli or nano:
799                expression.set("sec", sec)
800
801            return rename_func("MAKE_TIMESTAMP")(self, expression)
802
803        def tablesample_sql(
804            self,
805            expression: exp.TableSample,
806            tablesample_keyword: t.Optional[str] = None,
807        ) -> str:
808            if not isinstance(expression.parent, exp.Select):
809                # This sample clause only applies to a single source, not the entire resulting relation
810                tablesample_keyword = "TABLESAMPLE"
811
812            if expression.args.get("size"):
813                method = expression.args.get("method")
814                if method and method.name.upper() != "RESERVOIR":
815                    self.unsupported(
816                        f"Sampling method {method} is not supported with a discrete sample count, "
817                        "defaulting to reservoir sampling"
818                    )
819                    expression.set("method", exp.var("RESERVOIR"))
820
821            return super().tablesample_sql(expression, tablesample_keyword=tablesample_keyword)
822
823        def interval_sql(self, expression: exp.Interval) -> str:
824            multiplier: t.Optional[int] = None
825            unit = expression.text("unit").lower()
826
827            if unit.startswith("week"):
828                multiplier = 7
829            if unit.startswith("quarter"):
830                multiplier = 90
831
832            if multiplier:
833                return f"({multiplier} * {super().interval_sql(exp.Interval(this=expression.this, unit=exp.var('DAY')))})"
834
835            return super().interval_sql(expression)
836
837        def columndef_sql(self, expression: exp.ColumnDef, sep: str = " ") -> str:
838            if isinstance(expression.parent, exp.UserDefinedFunction):
839                return self.sql(expression, "this")
840            return super().columndef_sql(expression, sep)
841
842        def join_sql(self, expression: exp.Join) -> str:
843            if (
844                expression.side == "LEFT"
845                and not expression.args.get("on")
846                and isinstance(expression.this, exp.Unnest)
847            ):
848                # Some dialects support `LEFT JOIN UNNEST(...)` without an explicit ON clause
849                # DuckDB doesn't, but we can just add a dummy ON clause that is always true
850                return super().join_sql(expression.on(exp.true()))
851
852            return super().join_sql(expression)
853
854        def generateseries_sql(self, expression: exp.GenerateSeries) -> str:
855            # GENERATE_SERIES(a, b) -> [a, b], RANGE(a, b) -> [a, b)
856            if expression.args.get("is_end_exclusive"):
857                return rename_func("RANGE")(self, expression)
858
859            return self.function_fallback_sql(expression)
860
861        def bracket_sql(self, expression: exp.Bracket) -> str:
862            this = expression.this
863            if isinstance(this, exp.Array):
864                this.replace(exp.paren(this))
865
866            bracket = super().bracket_sql(expression)
867
868            if not expression.args.get("returns_list_for_maps"):
869                if not this.type:
870                    from sqlglot.optimizer.annotate_types import annotate_types
871
872                    this = annotate_types(this)
873
874                if this.is_type(exp.DataType.Type.MAP):
875                    bracket = f"({bracket})[1]"
876
877            return bracket
878
879        def withingroup_sql(self, expression: exp.WithinGroup) -> str:
880            expression_sql = self.sql(expression, "expression")
881
882            func = expression.this
883            if isinstance(func, exp.PERCENTILES):
884                # Make the order key the first arg and slide the fraction to the right
885                # https://duckdb.org/docs/sql/aggregates#ordered-set-aggregate-functions
886                order_col = expression.find(exp.Ordered)
887                if order_col:
888                    func.set("expression", func.this)
889                    func.set("this", order_col.this)
890
891            this = self.sql(expression, "this").rstrip(")")
892
893            return f"{this}{expression_sql})"
894
895        def length_sql(self, expression: exp.Length) -> str:
896            arg = expression.this
897
898            # Dialects like BQ and Snowflake also accept binary values as args, so
899            # DDB will attempt to infer the type or resort to case/when resolution
900            if not expression.args.get("binary") or arg.is_string:
901                return self.func("LENGTH", arg)
902
903            if not arg.type:
904                from sqlglot.optimizer.annotate_types import annotate_types
905
906                arg = annotate_types(arg)
907
908            if arg.is_type(*exp.DataType.TEXT_TYPES):
909                return self.func("LENGTH", arg)
910
911            # We need these casts to make duckdb's static type checker happy
912            blob = exp.cast(arg, exp.DataType.Type.VARBINARY)
913            varchar = exp.cast(arg, exp.DataType.Type.VARCHAR)
914
915            case = (
916                exp.case(self.func("TYPEOF", arg))
917                .when(
918                    "'VARCHAR'", exp.Anonymous(this="LENGTH", expressions=[varchar])
919                )  # anonymous to break length_sql recursion
920                .when("'BLOB'", self.func("OCTET_LENGTH", blob))
921            )
922
923            return self.sql(case)
924
925        def objectinsert_sql(self, expression: exp.ObjectInsert) -> str:
926            this = expression.this
927            key = expression.args.get("key")
928            key_sql = key.name if isinstance(key, exp.Expression) else ""
929            value_sql = self.sql(expression, "value")
930
931            kv_sql = f"{key_sql} := {value_sql}"
932
933            # If the input struct is empty e.g. transpiling OBJECT_INSERT(OBJECT_CONSTRUCT(), key, value) from Snowflake
934            # then we can generate STRUCT_PACK which will build it since STRUCT_INSERT({}, key := value) is not valid DuckDB
935            if isinstance(this, exp.Struct) and not this.expressions:
936                return self.func("STRUCT_PACK", kv_sql)
937
938            return self.func("STRUCT_INSERT", this, kv_sql)
939
940        def unnest_sql(self, expression: exp.Unnest) -> str:
941            explode_array = expression.args.get("explode_array")
942            if explode_array:
943                # In BigQuery, UNNESTing a nested array leads to explosion of the top-level array & struct
944                # This is transpiled to DDB by transforming "FROM UNNEST(...)" to "FROM (SELECT UNNEST(..., max_depth => 2))"
945                expression.expressions.append(
946                    exp.Kwarg(this=exp.var("max_depth"), expression=exp.Literal.number(2))
947                )
948
949                # If BQ's UNNEST is aliased, we transform it from a column alias to a table alias in DDB
950                alias = expression.args.get("alias")
951                if alias:
952                    expression.set("alias", None)
953                    alias = exp.TableAlias(this=seq_get(alias.args.get("columns"), 0))
954
955                unnest_sql = super().unnest_sql(expression)
956                select = exp.Select(expressions=[unnest_sql]).subquery(alias)
957                return self.sql(select)
958
959            return super().unnest_sql(expression)
960
961        def ignorenulls_sql(self, expression: exp.IgnoreNulls) -> str:
962            if isinstance(expression.this, WINDOW_FUNCS_WITH_IGNORE_NULLS):
963                # DuckDB should render IGNORE NULLS only for the general-purpose
964                # window functions that accept it e.g. FIRST_VALUE(... IGNORE NULLS) OVER (...)
965                return super().ignorenulls_sql(expression)
966
967            return self.sql(expression, "this")
968
969        def arraytostring_sql(self, expression: exp.ArrayToString) -> str:
970            this = self.sql(expression, "this")
971            null_text = self.sql(expression, "null")
972
973            if null_text:
974                this = f"LIST_TRANSFORM({this}, x -> COALESCE(x, {null_text}))"
975
976            return self.func("ARRAY_TO_STRING", this, expression.expression)
977
978        @unsupported_args("position", "occurrence")
979        def regexpextract_sql(self, expression: exp.RegexpExtract) -> str:
980            group = expression.args.get("group")
981            params = expression.args.get("parameters")
982
983            # Do not render group if there is no following argument,
984            # and it's the default value for this dialect
985            if (
986                not params
987                and group
988                and group.name == str(self.dialect.REGEXP_EXTRACT_DEFAULT_GROUP)
989            ):
990                group = None
991            return self.func(
992                "REGEXP_EXTRACT", expression.this, expression.expression, group, params
993            )
WINDOW_FUNCS_WITH_IGNORE_NULLS = (<class 'sqlglot.expressions.FirstValue'>, <class 'sqlglot.expressions.LastValue'>, <class 'sqlglot.expressions.Lag'>, <class 'sqlglot.expressions.Lead'>, <class 'sqlglot.expressions.NthValue'>)
WRAPPED_JSON_EXTRACT_EXPRESSIONS = (<class 'sqlglot.expressions.Binary'>, <class 'sqlglot.expressions.Bracket'>, <class 'sqlglot.expressions.In'>)
class DuckDB(sqlglot.dialects.dialect.Dialect):
275class DuckDB(Dialect):
276    NULL_ORDERING = "nulls_are_last"
277    SUPPORTS_USER_DEFINED_TYPES = False
278    SAFE_DIVISION = True
279    INDEX_OFFSET = 1
280    CONCAT_COALESCE = True
281    SUPPORTS_ORDER_BY_ALL = True
282    SUPPORTS_FIXED_SIZE_ARRAYS = True
283    STRICT_JSON_PATH_SYNTAX = False
284
285    # https://duckdb.org/docs/sql/introduction.html#creating-a-new-table
286    NORMALIZATION_STRATEGY = NormalizationStrategy.CASE_INSENSITIVE
287
288    def to_json_path(self, path: t.Optional[exp.Expression]) -> t.Optional[exp.Expression]:
289        if isinstance(path, exp.Literal):
290            # DuckDB also supports the JSON pointer syntax, where every path starts with a `/`.
291            # Additionally, it allows accessing the back of lists using the `[#-i]` syntax.
292            # This check ensures we'll avoid trying to parse these as JSON paths, which can
293            # either result in a noisy warning or in an invalid representation of the path.
294            path_text = path.name
295            if path_text.startswith("/") or "[#" in path_text:
296                return path
297
298        return super().to_json_path(path)
299
300    class Tokenizer(tokens.Tokenizer):
301        HEREDOC_STRINGS = ["$"]
302
303        HEREDOC_TAG_IS_IDENTIFIER = True
304        HEREDOC_STRING_ALTERNATIVE = TokenType.PARAMETER
305
306        KEYWORDS = {
307            **tokens.Tokenizer.KEYWORDS,
308            "//": TokenType.DIV,
309            "**": TokenType.DSTAR,
310            "^@": TokenType.CARET_AT,
311            "@>": TokenType.AT_GT,
312            "<@": TokenType.LT_AT,
313            "ATTACH": TokenType.COMMAND,
314            "BINARY": TokenType.VARBINARY,
315            "BITSTRING": TokenType.BIT,
316            "BPCHAR": TokenType.TEXT,
317            "CHAR": TokenType.TEXT,
318            "CHARACTER VARYING": TokenType.TEXT,
319            "EXCLUDE": TokenType.EXCEPT,
320            "LOGICAL": TokenType.BOOLEAN,
321            "ONLY": TokenType.ONLY,
322            "PIVOT_WIDER": TokenType.PIVOT,
323            "POSITIONAL": TokenType.POSITIONAL,
324            "SIGNED": TokenType.INT,
325            "STRING": TokenType.TEXT,
326            "SUMMARIZE": TokenType.SUMMARIZE,
327            "TIMESTAMP_S": TokenType.TIMESTAMP_S,
328            "TIMESTAMP_MS": TokenType.TIMESTAMP_MS,
329            "TIMESTAMP_NS": TokenType.TIMESTAMP_NS,
330            "TIMESTAMP_US": TokenType.TIMESTAMP,
331            "UBIGINT": TokenType.UBIGINT,
332            "UINTEGER": TokenType.UINT,
333            "USMALLINT": TokenType.USMALLINT,
334            "UTINYINT": TokenType.UTINYINT,
335            "VARCHAR": TokenType.TEXT,
336        }
337        KEYWORDS.pop("/*+")
338
339        SINGLE_TOKENS = {
340            **tokens.Tokenizer.SINGLE_TOKENS,
341            "$": TokenType.PARAMETER,
342        }
343
344    class Parser(parser.Parser):
345        BITWISE = {
346            **parser.Parser.BITWISE,
347            TokenType.TILDA: exp.RegexpLike,
348        }
349        BITWISE.pop(TokenType.CARET)
350
351        RANGE_PARSERS = {
352            **parser.Parser.RANGE_PARSERS,
353            TokenType.DAMP: binary_range_parser(exp.ArrayOverlaps),
354            TokenType.CARET_AT: binary_range_parser(exp.StartsWith),
355        }
356
357        EXPONENT = {
358            **parser.Parser.EXPONENT,
359            TokenType.CARET: exp.Pow,
360            TokenType.DSTAR: exp.Pow,
361        }
362
363        FUNCTIONS_WITH_ALIASED_ARGS = {*parser.Parser.FUNCTIONS_WITH_ALIASED_ARGS, "STRUCT_PACK"}
364
365        FUNCTIONS = {
366            **parser.Parser.FUNCTIONS,
367            "ARRAY_REVERSE_SORT": _build_sort_array_desc,
368            "ARRAY_SORT": exp.SortArray.from_arg_list,
369            "DATEDIFF": _build_date_diff,
370            "DATE_DIFF": _build_date_diff,
371            "DATE_TRUNC": date_trunc_to_time,
372            "DATETRUNC": date_trunc_to_time,
373            "DECODE": lambda args: exp.Decode(
374                this=seq_get(args, 0), charset=exp.Literal.string("utf-8")
375            ),
376            "ENCODE": lambda args: exp.Encode(
377                this=seq_get(args, 0), charset=exp.Literal.string("utf-8")
378            ),
379            "EPOCH": exp.TimeToUnix.from_arg_list,
380            "EPOCH_MS": lambda args: exp.UnixToTime(
381                this=seq_get(args, 0), scale=exp.UnixToTime.MILLIS
382            ),
383            "JSON": exp.ParseJSON.from_arg_list,
384            "JSON_EXTRACT_PATH": parser.build_extract_json_with_path(exp.JSONExtract),
385            "JSON_EXTRACT_STRING": parser.build_extract_json_with_path(exp.JSONExtractScalar),
386            "LIST_HAS": exp.ArrayContains.from_arg_list,
387            "LIST_REVERSE_SORT": _build_sort_array_desc,
388            "LIST_SORT": exp.SortArray.from_arg_list,
389            "LIST_VALUE": lambda args: exp.Array(expressions=args),
390            "MAKE_TIME": exp.TimeFromParts.from_arg_list,
391            "MAKE_TIMESTAMP": _build_make_timestamp,
392            "QUANTILE_CONT": exp.PercentileCont.from_arg_list,
393            "QUANTILE_DISC": exp.PercentileDisc.from_arg_list,
394            "REGEXP_EXTRACT": build_regexp_extract(exp.RegexpExtract),
395            "REGEXP_EXTRACT_ALL": build_regexp_extract(exp.RegexpExtractAll),
396            "REGEXP_MATCHES": exp.RegexpLike.from_arg_list,
397            "REGEXP_REPLACE": lambda args: exp.RegexpReplace(
398                this=seq_get(args, 0),
399                expression=seq_get(args, 1),
400                replacement=seq_get(args, 2),
401                modifiers=seq_get(args, 3),
402            ),
403            "STRFTIME": build_formatted_time(exp.TimeToStr, "duckdb"),
404            "STRING_SPLIT": exp.Split.from_arg_list,
405            "STRING_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
406            "STRING_TO_ARRAY": exp.Split.from_arg_list,
407            "STRPTIME": build_formatted_time(exp.StrToTime, "duckdb"),
408            "STRUCT_PACK": exp.Struct.from_arg_list,
409            "STR_SPLIT": exp.Split.from_arg_list,
410            "STR_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
411            "TO_TIMESTAMP": exp.UnixToTime.from_arg_list,
412            "UNNEST": exp.Explode.from_arg_list,
413            "XOR": binary_from_function(exp.BitwiseXor),
414            "GENERATE_SERIES": _build_generate_series(),
415            "RANGE": _build_generate_series(end_exclusive=True),
416            "EDITDIST3": exp.Levenshtein.from_arg_list,
417        }
418
419        FUNCTIONS.pop("DATE_SUB")
420        FUNCTIONS.pop("GLOB")
421
422        FUNCTION_PARSERS = parser.Parser.FUNCTION_PARSERS.copy()
423        FUNCTION_PARSERS.pop("DECODE")
424
425        NO_PAREN_FUNCTION_PARSERS = {
426            **parser.Parser.NO_PAREN_FUNCTION_PARSERS,
427            "MAP": lambda self: self._parse_map(),
428        }
429
430        TABLE_ALIAS_TOKENS = parser.Parser.TABLE_ALIAS_TOKENS - {
431            TokenType.SEMI,
432            TokenType.ANTI,
433        }
434
435        PLACEHOLDER_PARSERS = {
436            **parser.Parser.PLACEHOLDER_PARSERS,
437            TokenType.PARAMETER: lambda self: (
438                self.expression(exp.Placeholder, this=self._prev.text)
439                if self._match(TokenType.NUMBER) or self._match_set(self.ID_VAR_TOKENS)
440                else None
441            ),
442        }
443
444        TYPE_CONVERTERS = {
445            # https://duckdb.org/docs/sql/data_types/numeric
446            exp.DataType.Type.DECIMAL: build_default_decimal_type(precision=18, scale=3),
447            # https://duckdb.org/docs/sql/data_types/text
448            exp.DataType.Type.TEXT: lambda dtype: exp.DataType.build("TEXT"),
449        }
450
451        def _parse_table_sample(self, as_modifier: bool = False) -> t.Optional[exp.TableSample]:
452            # https://duckdb.org/docs/sql/samples.html
453            sample = super()._parse_table_sample(as_modifier=as_modifier)
454            if sample and not sample.args.get("method"):
455                if sample.args.get("size"):
456                    sample.set("method", exp.var("RESERVOIR"))
457                else:
458                    sample.set("method", exp.var("SYSTEM"))
459
460            return sample
461
462        def _parse_bracket(
463            self, this: t.Optional[exp.Expression] = None
464        ) -> t.Optional[exp.Expression]:
465            bracket = super()._parse_bracket(this)
466            if isinstance(bracket, exp.Bracket):
467                bracket.set("returns_list_for_maps", True)
468
469            return bracket
470
471        def _parse_map(self) -> exp.ToMap | exp.Map:
472            if self._match(TokenType.L_BRACE, advance=False):
473                return self.expression(exp.ToMap, this=self._parse_bracket())
474
475            args = self._parse_wrapped_csv(self._parse_assignment)
476            return self.expression(exp.Map, keys=seq_get(args, 0), values=seq_get(args, 1))
477
478        def _parse_struct_types(self, type_required: bool = False) -> t.Optional[exp.Expression]:
479            return self._parse_field_def()
480
481        def _pivot_column_names(self, aggregations: t.List[exp.Expression]) -> t.List[str]:
482            if len(aggregations) == 1:
483                return super()._pivot_column_names(aggregations)
484            return pivot_column_names(aggregations, dialect="duckdb")
485
486    class Generator(generator.Generator):
487        PARAMETER_TOKEN = "$"
488        NAMED_PLACEHOLDER_TOKEN = "$"
489        JOIN_HINTS = False
490        TABLE_HINTS = False
491        QUERY_HINTS = False
492        LIMIT_FETCH = "LIMIT"
493        STRUCT_DELIMITER = ("(", ")")
494        RENAME_TABLE_WITH_DB = False
495        NVL2_SUPPORTED = False
496        SEMI_ANTI_JOIN_WITH_SIDE = False
497        TABLESAMPLE_KEYWORDS = "USING SAMPLE"
498        TABLESAMPLE_SEED_KEYWORD = "REPEATABLE"
499        LAST_DAY_SUPPORTS_DATE_PART = False
500        JSON_KEY_VALUE_PAIR_SEP = ","
501        IGNORE_NULLS_IN_FUNC = True
502        JSON_PATH_BRACKETED_KEY_SUPPORTED = False
503        SUPPORTS_CREATE_TABLE_LIKE = False
504        MULTI_ARG_DISTINCT = False
505        CAN_IMPLEMENT_ARRAY_ANY = True
506        SUPPORTS_TO_NUMBER = False
507        COPY_HAS_INTO_KEYWORD = False
508        STAR_EXCEPT = "EXCLUDE"
509        PAD_FILL_PATTERN_IS_REQUIRED = True
510        ARRAY_CONCAT_IS_VAR_LEN = False
511        ARRAY_SIZE_DIM_REQUIRED = False
512
513        TRANSFORMS = {
514            **generator.Generator.TRANSFORMS,
515            exp.ApproxDistinct: approx_count_distinct_sql,
516            exp.Array: inline_array_unless_query,
517            exp.ArrayFilter: rename_func("LIST_FILTER"),
518            exp.ArgMax: arg_max_or_min_no_count("ARG_MAX"),
519            exp.ArgMin: arg_max_or_min_no_count("ARG_MIN"),
520            exp.ArraySort: _array_sort_sql,
521            exp.ArraySum: rename_func("LIST_SUM"),
522            exp.BitwiseXor: rename_func("XOR"),
523            exp.CommentColumnConstraint: no_comment_column_constraint_sql,
524            exp.CurrentDate: lambda *_: "CURRENT_DATE",
525            exp.CurrentTime: lambda *_: "CURRENT_TIME",
526            exp.CurrentTimestamp: lambda *_: "CURRENT_TIMESTAMP",
527            exp.DayOfMonth: rename_func("DAYOFMONTH"),
528            exp.DayOfWeek: rename_func("DAYOFWEEK"),
529            exp.DayOfWeekIso: rename_func("ISODOW"),
530            exp.DayOfYear: rename_func("DAYOFYEAR"),
531            exp.DataType: _datatype_sql,
532            exp.Date: _date_sql,
533            exp.DateAdd: _date_delta_sql,
534            exp.DateFromParts: rename_func("MAKE_DATE"),
535            exp.DateSub: _date_delta_sql,
536            exp.DateDiff: _date_diff_sql,
537            exp.DateStrToDate: datestrtodate_sql,
538            exp.Datetime: no_datetime_sql,
539            exp.DatetimeSub: _date_delta_sql,
540            exp.DatetimeAdd: _date_delta_sql,
541            exp.DateToDi: lambda self,
542            e: f"CAST(STRFTIME({self.sql(e, 'this')}, {DuckDB.DATEINT_FORMAT}) AS INT)",
543            exp.Decode: lambda self, e: encode_decode_sql(self, e, "DECODE", replace=False),
544            exp.DiToDate: lambda self,
545            e: f"CAST(STRPTIME(CAST({self.sql(e, 'this')} AS TEXT), {DuckDB.DATEINT_FORMAT}) AS DATE)",
546            exp.Encode: lambda self, e: encode_decode_sql(self, e, "ENCODE", replace=False),
547            exp.GenerateDateArray: _generate_datetime_array_sql,
548            exp.GenerateTimestampArray: _generate_datetime_array_sql,
549            exp.Explode: rename_func("UNNEST"),
550            exp.IntDiv: lambda self, e: self.binary(e, "//"),
551            exp.IsInf: rename_func("ISINF"),
552            exp.IsNan: rename_func("ISNAN"),
553            exp.JSONBExists: rename_func("JSON_EXISTS"),
554            exp.JSONExtract: _arrow_json_extract_sql,
555            exp.JSONExtractArray: _json_extract_value_array_sql,
556            exp.JSONExtractScalar: _arrow_json_extract_sql,
557            exp.JSONFormat: _json_format_sql,
558            exp.JSONValueArray: _json_extract_value_array_sql,
559            exp.Lateral: explode_to_unnest_sql,
560            exp.LogicalOr: rename_func("BOOL_OR"),
561            exp.LogicalAnd: rename_func("BOOL_AND"),
562            exp.MD5Digest: lambda self, e: self.func("UNHEX", self.func("MD5", e.this)),
563            exp.MonthsBetween: lambda self, e: self.func(
564                "DATEDIFF",
565                "'month'",
566                exp.cast(e.expression, exp.DataType.Type.TIMESTAMP, copy=True),
567                exp.cast(e.this, exp.DataType.Type.TIMESTAMP, copy=True),
568            ),
569            exp.PercentileCont: rename_func("QUANTILE_CONT"),
570            exp.PercentileDisc: rename_func("QUANTILE_DISC"),
571            # DuckDB doesn't allow qualified columns inside of PIVOT expressions.
572            # See: https://github.com/duckdb/duckdb/blob/671faf92411182f81dce42ac43de8bfb05d9909e/src/planner/binder/tableref/bind_pivot.cpp#L61-L62
573            exp.Pivot: transforms.preprocess([transforms.unqualify_columns]),
574            exp.RegexpReplace: lambda self, e: self.func(
575                "REGEXP_REPLACE",
576                e.this,
577                e.expression,
578                e.args.get("replacement"),
579                e.args.get("modifiers"),
580            ),
581            exp.RegexpLike: rename_func("REGEXP_MATCHES"),
582            exp.RegexpSplit: rename_func("STR_SPLIT_REGEX"),
583            exp.Return: lambda self, e: self.sql(e, "this"),
584            exp.ReturnsProperty: lambda self, e: "TABLE" if isinstance(e.this, exp.Schema) else "",
585            exp.Rand: rename_func("RANDOM"),
586            exp.SafeDivide: no_safe_divide_sql,
587            exp.SHA: rename_func("SHA1"),
588            exp.SHA2: sha256_sql,
589            exp.Split: rename_func("STR_SPLIT"),
590            exp.SortArray: _sort_array_sql,
591            exp.StrPosition: str_position_sql,
592            exp.StrToUnix: lambda self, e: self.func(
593                "EPOCH", self.func("STRPTIME", e.this, self.format_time(e))
594            ),
595            exp.Struct: _struct_sql,
596            exp.Transform: rename_func("LIST_TRANSFORM"),
597            exp.TimeAdd: _date_delta_sql,
598            exp.Time: no_time_sql,
599            exp.TimeDiff: _timediff_sql,
600            exp.Timestamp: no_timestamp_sql,
601            exp.TimestampDiff: lambda self, e: self.func(
602                "DATE_DIFF", exp.Literal.string(e.unit), e.expression, e.this
603            ),
604            exp.TimestampTrunc: timestamptrunc_sql(),
605            exp.TimeStrToDate: lambda self, e: self.sql(exp.cast(e.this, exp.DataType.Type.DATE)),
606            exp.TimeStrToTime: timestrtotime_sql,
607            exp.TimeStrToUnix: lambda self, e: self.func(
608                "EPOCH", exp.cast(e.this, exp.DataType.Type.TIMESTAMP)
609            ),
610            exp.TimeToStr: lambda self, e: self.func("STRFTIME", e.this, self.format_time(e)),
611            exp.TimeToUnix: rename_func("EPOCH"),
612            exp.TsOrDiToDi: lambda self,
613            e: f"CAST(SUBSTR(REPLACE(CAST({self.sql(e, 'this')} AS TEXT), '-', ''), 1, 8) AS INT)",
614            exp.TsOrDsAdd: _date_delta_sql,
615            exp.TsOrDsDiff: lambda self, e: self.func(
616                "DATE_DIFF",
617                f"'{e.args.get('unit') or 'DAY'}'",
618                exp.cast(e.expression, exp.DataType.Type.TIMESTAMP),
619                exp.cast(e.this, exp.DataType.Type.TIMESTAMP),
620            ),
621            exp.UnixToStr: lambda self, e: self.func(
622                "STRFTIME", self.func("TO_TIMESTAMP", e.this), self.format_time(e)
623            ),
624            exp.DatetimeTrunc: lambda self, e: self.func(
625                "DATE_TRUNC", unit_to_str(e), exp.cast(e.this, exp.DataType.Type.DATETIME)
626            ),
627            exp.UnixToTime: _unix_to_time_sql,
628            exp.UnixToTimeStr: lambda self, e: f"CAST(TO_TIMESTAMP({self.sql(e, 'this')}) AS TEXT)",
629            exp.VariancePop: rename_func("VAR_POP"),
630            exp.WeekOfYear: rename_func("WEEKOFYEAR"),
631            exp.Xor: bool_xor_sql,
632            exp.Levenshtein: unsupported_args("ins_cost", "del_cost", "sub_cost", "max_dist")(
633                rename_func("LEVENSHTEIN")
634            ),
635        }
636
637        SUPPORTED_JSON_PATH_PARTS = {
638            exp.JSONPathKey,
639            exp.JSONPathRoot,
640            exp.JSONPathSubscript,
641            exp.JSONPathWildcard,
642        }
643
644        TYPE_MAPPING = {
645            **generator.Generator.TYPE_MAPPING,
646            exp.DataType.Type.BINARY: "BLOB",
647            exp.DataType.Type.BPCHAR: "TEXT",
648            exp.DataType.Type.CHAR: "TEXT",
649            exp.DataType.Type.FLOAT: "REAL",
650            exp.DataType.Type.NCHAR: "TEXT",
651            exp.DataType.Type.NVARCHAR: "TEXT",
652            exp.DataType.Type.UINT: "UINTEGER",
653            exp.DataType.Type.VARBINARY: "BLOB",
654            exp.DataType.Type.ROWVERSION: "BLOB",
655            exp.DataType.Type.VARCHAR: "TEXT",
656            exp.DataType.Type.TIMESTAMPNTZ: "TIMESTAMP",
657            exp.DataType.Type.TIMESTAMP_S: "TIMESTAMP_S",
658            exp.DataType.Type.TIMESTAMP_MS: "TIMESTAMP_MS",
659            exp.DataType.Type.TIMESTAMP_NS: "TIMESTAMP_NS",
660        }
661
662        # https://github.com/duckdb/duckdb/blob/ff7f24fd8e3128d94371827523dae85ebaf58713/third_party/libpg_query/grammar/keywords/reserved_keywords.list#L1-L77
663        RESERVED_KEYWORDS = {
664            "array",
665            "analyse",
666            "union",
667            "all",
668            "when",
669            "in_p",
670            "default",
671            "create_p",
672            "window",
673            "asymmetric",
674            "to",
675            "else",
676            "localtime",
677            "from",
678            "end_p",
679            "select",
680            "current_date",
681            "foreign",
682            "with",
683            "grant",
684            "session_user",
685            "or",
686            "except",
687            "references",
688            "fetch",
689            "limit",
690            "group_p",
691            "leading",
692            "into",
693            "collate",
694            "offset",
695            "do",
696            "then",
697            "localtimestamp",
698            "check_p",
699            "lateral_p",
700            "current_role",
701            "where",
702            "asc_p",
703            "placing",
704            "desc_p",
705            "user",
706            "unique",
707            "initially",
708            "column",
709            "both",
710            "some",
711            "as",
712            "any",
713            "only",
714            "deferrable",
715            "null_p",
716            "current_time",
717            "true_p",
718            "table",
719            "case",
720            "trailing",
721            "variadic",
722            "for",
723            "on",
724            "distinct",
725            "false_p",
726            "not",
727            "constraint",
728            "current_timestamp",
729            "returning",
730            "primary",
731            "intersect",
732            "having",
733            "analyze",
734            "current_user",
735            "and",
736            "cast",
737            "symmetric",
738            "using",
739            "order",
740            "current_catalog",
741        }
742
743        UNWRAPPED_INTERVAL_VALUES = (exp.Literal, exp.Paren)
744
745        # DuckDB doesn't generally support CREATE TABLE .. properties
746        # https://duckdb.org/docs/sql/statements/create_table.html
747        PROPERTIES_LOCATION = {
748            prop: exp.Properties.Location.UNSUPPORTED
749            for prop in generator.Generator.PROPERTIES_LOCATION
750        }
751
752        # There are a few exceptions (e.g. temporary tables) which are supported or
753        # can be transpiled to DuckDB, so we explicitly override them accordingly
754        PROPERTIES_LOCATION[exp.LikeProperty] = exp.Properties.Location.POST_SCHEMA
755        PROPERTIES_LOCATION[exp.TemporaryProperty] = exp.Properties.Location.POST_CREATE
756        PROPERTIES_LOCATION[exp.ReturnsProperty] = exp.Properties.Location.POST_ALIAS
757
758        def fromiso8601timestamp_sql(self, expression: exp.FromISO8601Timestamp) -> str:
759            return self.sql(exp.cast(expression.this, exp.DataType.Type.TIMESTAMPTZ))
760
761        def strtotime_sql(self, expression: exp.StrToTime) -> str:
762            if expression.args.get("safe"):
763                formatted_time = self.format_time(expression)
764                return f"CAST({self.func('TRY_STRPTIME', expression.this, formatted_time)} AS TIMESTAMP)"
765            return str_to_time_sql(self, expression)
766
767        def strtodate_sql(self, expression: exp.StrToDate) -> str:
768            if expression.args.get("safe"):
769                formatted_time = self.format_time(expression)
770                return f"CAST({self.func('TRY_STRPTIME', expression.this, formatted_time)} AS DATE)"
771            return f"CAST({str_to_time_sql(self, expression)} AS DATE)"
772
773        def parsejson_sql(self, expression: exp.ParseJSON) -> str:
774            arg = expression.this
775            if expression.args.get("safe"):
776                return self.sql(exp.case().when(exp.func("json_valid", arg), arg).else_(exp.null()))
777            return self.func("JSON", arg)
778
779        def timefromparts_sql(self, expression: exp.TimeFromParts) -> str:
780            nano = expression.args.get("nano")
781            if nano is not None:
782                expression.set(
783                    "sec", expression.args["sec"] + nano.pop() / exp.Literal.number(1000000000.0)
784                )
785
786            return rename_func("MAKE_TIME")(self, expression)
787
788        def timestampfromparts_sql(self, expression: exp.TimestampFromParts) -> str:
789            sec = expression.args["sec"]
790
791            milli = expression.args.get("milli")
792            if milli is not None:
793                sec += milli.pop() / exp.Literal.number(1000.0)
794
795            nano = expression.args.get("nano")
796            if nano is not None:
797                sec += nano.pop() / exp.Literal.number(1000000000.0)
798
799            if milli or nano:
800                expression.set("sec", sec)
801
802            return rename_func("MAKE_TIMESTAMP")(self, expression)
803
804        def tablesample_sql(
805            self,
806            expression: exp.TableSample,
807            tablesample_keyword: t.Optional[str] = None,
808        ) -> str:
809            if not isinstance(expression.parent, exp.Select):
810                # This sample clause only applies to a single source, not the entire resulting relation
811                tablesample_keyword = "TABLESAMPLE"
812
813            if expression.args.get("size"):
814                method = expression.args.get("method")
815                if method and method.name.upper() != "RESERVOIR":
816                    self.unsupported(
817                        f"Sampling method {method} is not supported with a discrete sample count, "
818                        "defaulting to reservoir sampling"
819                    )
820                    expression.set("method", exp.var("RESERVOIR"))
821
822            return super().tablesample_sql(expression, tablesample_keyword=tablesample_keyword)
823
824        def interval_sql(self, expression: exp.Interval) -> str:
825            multiplier: t.Optional[int] = None
826            unit = expression.text("unit").lower()
827
828            if unit.startswith("week"):
829                multiplier = 7
830            if unit.startswith("quarter"):
831                multiplier = 90
832
833            if multiplier:
834                return f"({multiplier} * {super().interval_sql(exp.Interval(this=expression.this, unit=exp.var('DAY')))})"
835
836            return super().interval_sql(expression)
837
838        def columndef_sql(self, expression: exp.ColumnDef, sep: str = " ") -> str:
839            if isinstance(expression.parent, exp.UserDefinedFunction):
840                return self.sql(expression, "this")
841            return super().columndef_sql(expression, sep)
842
843        def join_sql(self, expression: exp.Join) -> str:
844            if (
845                expression.side == "LEFT"
846                and not expression.args.get("on")
847                and isinstance(expression.this, exp.Unnest)
848            ):
849                # Some dialects support `LEFT JOIN UNNEST(...)` without an explicit ON clause
850                # DuckDB doesn't, but we can just add a dummy ON clause that is always true
851                return super().join_sql(expression.on(exp.true()))
852
853            return super().join_sql(expression)
854
855        def generateseries_sql(self, expression: exp.GenerateSeries) -> str:
856            # GENERATE_SERIES(a, b) -> [a, b], RANGE(a, b) -> [a, b)
857            if expression.args.get("is_end_exclusive"):
858                return rename_func("RANGE")(self, expression)
859
860            return self.function_fallback_sql(expression)
861
862        def bracket_sql(self, expression: exp.Bracket) -> str:
863            this = expression.this
864            if isinstance(this, exp.Array):
865                this.replace(exp.paren(this))
866
867            bracket = super().bracket_sql(expression)
868
869            if not expression.args.get("returns_list_for_maps"):
870                if not this.type:
871                    from sqlglot.optimizer.annotate_types import annotate_types
872
873                    this = annotate_types(this)
874
875                if this.is_type(exp.DataType.Type.MAP):
876                    bracket = f"({bracket})[1]"
877
878            return bracket
879
880        def withingroup_sql(self, expression: exp.WithinGroup) -> str:
881            expression_sql = self.sql(expression, "expression")
882
883            func = expression.this
884            if isinstance(func, exp.PERCENTILES):
885                # Make the order key the first arg and slide the fraction to the right
886                # https://duckdb.org/docs/sql/aggregates#ordered-set-aggregate-functions
887                order_col = expression.find(exp.Ordered)
888                if order_col:
889                    func.set("expression", func.this)
890                    func.set("this", order_col.this)
891
892            this = self.sql(expression, "this").rstrip(")")
893
894            return f"{this}{expression_sql})"
895
896        def length_sql(self, expression: exp.Length) -> str:
897            arg = expression.this
898
899            # Dialects like BQ and Snowflake also accept binary values as args, so
900            # DDB will attempt to infer the type or resort to case/when resolution
901            if not expression.args.get("binary") or arg.is_string:
902                return self.func("LENGTH", arg)
903
904            if not arg.type:
905                from sqlglot.optimizer.annotate_types import annotate_types
906
907                arg = annotate_types(arg)
908
909            if arg.is_type(*exp.DataType.TEXT_TYPES):
910                return self.func("LENGTH", arg)
911
912            # We need these casts to make duckdb's static type checker happy
913            blob = exp.cast(arg, exp.DataType.Type.VARBINARY)
914            varchar = exp.cast(arg, exp.DataType.Type.VARCHAR)
915
916            case = (
917                exp.case(self.func("TYPEOF", arg))
918                .when(
919                    "'VARCHAR'", exp.Anonymous(this="LENGTH", expressions=[varchar])
920                )  # anonymous to break length_sql recursion
921                .when("'BLOB'", self.func("OCTET_LENGTH", blob))
922            )
923
924            return self.sql(case)
925
926        def objectinsert_sql(self, expression: exp.ObjectInsert) -> str:
927            this = expression.this
928            key = expression.args.get("key")
929            key_sql = key.name if isinstance(key, exp.Expression) else ""
930            value_sql = self.sql(expression, "value")
931
932            kv_sql = f"{key_sql} := {value_sql}"
933
934            # If the input struct is empty e.g. transpiling OBJECT_INSERT(OBJECT_CONSTRUCT(), key, value) from Snowflake
935            # then we can generate STRUCT_PACK which will build it since STRUCT_INSERT({}, key := value) is not valid DuckDB
936            if isinstance(this, exp.Struct) and not this.expressions:
937                return self.func("STRUCT_PACK", kv_sql)
938
939            return self.func("STRUCT_INSERT", this, kv_sql)
940
941        def unnest_sql(self, expression: exp.Unnest) -> str:
942            explode_array = expression.args.get("explode_array")
943            if explode_array:
944                # In BigQuery, UNNESTing a nested array leads to explosion of the top-level array & struct
945                # This is transpiled to DDB by transforming "FROM UNNEST(...)" to "FROM (SELECT UNNEST(..., max_depth => 2))"
946                expression.expressions.append(
947                    exp.Kwarg(this=exp.var("max_depth"), expression=exp.Literal.number(2))
948                )
949
950                # If BQ's UNNEST is aliased, we transform it from a column alias to a table alias in DDB
951                alias = expression.args.get("alias")
952                if alias:
953                    expression.set("alias", None)
954                    alias = exp.TableAlias(this=seq_get(alias.args.get("columns"), 0))
955
956                unnest_sql = super().unnest_sql(expression)
957                select = exp.Select(expressions=[unnest_sql]).subquery(alias)
958                return self.sql(select)
959
960            return super().unnest_sql(expression)
961
962        def ignorenulls_sql(self, expression: exp.IgnoreNulls) -> str:
963            if isinstance(expression.this, WINDOW_FUNCS_WITH_IGNORE_NULLS):
964                # DuckDB should render IGNORE NULLS only for the general-purpose
965                # window functions that accept it e.g. FIRST_VALUE(... IGNORE NULLS) OVER (...)
966                return super().ignorenulls_sql(expression)
967
968            return self.sql(expression, "this")
969
970        def arraytostring_sql(self, expression: exp.ArrayToString) -> str:
971            this = self.sql(expression, "this")
972            null_text = self.sql(expression, "null")
973
974            if null_text:
975                this = f"LIST_TRANSFORM({this}, x -> COALESCE(x, {null_text}))"
976
977            return self.func("ARRAY_TO_STRING", this, expression.expression)
978
979        @unsupported_args("position", "occurrence")
980        def regexpextract_sql(self, expression: exp.RegexpExtract) -> str:
981            group = expression.args.get("group")
982            params = expression.args.get("parameters")
983
984            # Do not render group if there is no following argument,
985            # and it's the default value for this dialect
986            if (
987                not params
988                and group
989                and group.name == str(self.dialect.REGEXP_EXTRACT_DEFAULT_GROUP)
990            ):
991                group = None
992            return self.func(
993                "REGEXP_EXTRACT", expression.this, expression.expression, group, params
994            )
NULL_ORDERING = 'nulls_are_last'

Default NULL ordering method to use if not explicitly set. Possible values: "nulls_are_small", "nulls_are_large", "nulls_are_last"

SUPPORTS_USER_DEFINED_TYPES = False

Whether user-defined data types are supported.

SAFE_DIVISION = True

Whether division by zero throws an error (False) or returns NULL (True).

INDEX_OFFSET = 1

The base index offset for arrays.

CONCAT_COALESCE = True

A NULL arg in CONCAT yields NULL by default, but in some dialects it yields an empty string.

SUPPORTS_ORDER_BY_ALL = True

Whether ORDER BY ALL is supported (expands to all the selected columns) as in DuckDB, Spark3/Databricks

SUPPORTS_FIXED_SIZE_ARRAYS = True

Whether expressions such as x::INT[5] should be parsed as fixed-size array defs/casts e.g. in DuckDB. In dialects which don't support fixed size arrays such as Snowflake, this should be interpreted as a subscript/index operator.

STRICT_JSON_PATH_SYNTAX = False

Whether failing to parse a JSON path expression using the JSONPath dialect will log a warning.

NORMALIZATION_STRATEGY = <NormalizationStrategy.CASE_INSENSITIVE: 'CASE_INSENSITIVE'>

Specifies the strategy according to which identifiers should be normalized.

def to_json_path( self, path: Optional[sqlglot.expressions.Expression]) -> Optional[sqlglot.expressions.Expression]:
288    def to_json_path(self, path: t.Optional[exp.Expression]) -> t.Optional[exp.Expression]:
289        if isinstance(path, exp.Literal):
290            # DuckDB also supports the JSON pointer syntax, where every path starts with a `/`.
291            # Additionally, it allows accessing the back of lists using the `[#-i]` syntax.
292            # This check ensures we'll avoid trying to parse these as JSON paths, which can
293            # either result in a noisy warning or in an invalid representation of the path.
294            path_text = path.name
295            if path_text.startswith("/") or "[#" in path_text:
296                return path
297
298        return super().to_json_path(path)
SUPPORTS_COLUMN_JOIN_MARKS = False

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

tokenizer_class = <class 'DuckDB.Tokenizer'>
jsonpath_tokenizer_class = <class 'sqlglot.tokens.JSONPathTokenizer'>
parser_class = <class 'DuckDB.Parser'>
generator_class = <class 'DuckDB.Generator'>
TIME_TRIE: Dict = {}
FORMAT_TRIE: Dict = {}
INVERSE_TIME_MAPPING: Dict[str, str] = {}
INVERSE_TIME_TRIE: Dict = {}
INVERSE_FORMAT_MAPPING: Dict[str, str] = {}
INVERSE_FORMAT_TRIE: Dict = {}
INVERSE_CREATABLE_KIND_MAPPING: dict[str, str] = {}
ESCAPED_SEQUENCES: Dict[str, str] = {}
QUOTE_START = "'"
QUOTE_END = "'"
IDENTIFIER_START = '"'
IDENTIFIER_END = '"'
BIT_START: Optional[str] = None
BIT_END: Optional[str] = None
HEX_START: Optional[str] = None
HEX_END: Optional[str] = None
BYTE_START: Optional[str] = None
BYTE_END: Optional[str] = None
UNICODE_START: Optional[str] = None
UNICODE_END: Optional[str] = None
class DuckDB.Tokenizer(sqlglot.tokens.Tokenizer):
300    class Tokenizer(tokens.Tokenizer):
301        HEREDOC_STRINGS = ["$"]
302
303        HEREDOC_TAG_IS_IDENTIFIER = True
304        HEREDOC_STRING_ALTERNATIVE = TokenType.PARAMETER
305
306        KEYWORDS = {
307            **tokens.Tokenizer.KEYWORDS,
308            "//": TokenType.DIV,
309            "**": TokenType.DSTAR,
310            "^@": TokenType.CARET_AT,
311            "@>": TokenType.AT_GT,
312            "<@": TokenType.LT_AT,
313            "ATTACH": TokenType.COMMAND,
314            "BINARY": TokenType.VARBINARY,
315            "BITSTRING": TokenType.BIT,
316            "BPCHAR": TokenType.TEXT,
317            "CHAR": TokenType.TEXT,
318            "CHARACTER VARYING": TokenType.TEXT,
319            "EXCLUDE": TokenType.EXCEPT,
320            "LOGICAL": TokenType.BOOLEAN,
321            "ONLY": TokenType.ONLY,
322            "PIVOT_WIDER": TokenType.PIVOT,
323            "POSITIONAL": TokenType.POSITIONAL,
324            "SIGNED": TokenType.INT,
325            "STRING": TokenType.TEXT,
326            "SUMMARIZE": TokenType.SUMMARIZE,
327            "TIMESTAMP_S": TokenType.TIMESTAMP_S,
328            "TIMESTAMP_MS": TokenType.TIMESTAMP_MS,
329            "TIMESTAMP_NS": TokenType.TIMESTAMP_NS,
330            "TIMESTAMP_US": TokenType.TIMESTAMP,
331            "UBIGINT": TokenType.UBIGINT,
332            "UINTEGER": TokenType.UINT,
333            "USMALLINT": TokenType.USMALLINT,
334            "UTINYINT": TokenType.UTINYINT,
335            "VARCHAR": TokenType.TEXT,
336        }
337        KEYWORDS.pop("/*+")
338
339        SINGLE_TOKENS = {
340            **tokens.Tokenizer.SINGLE_TOKENS,
341            "$": TokenType.PARAMETER,
342        }
HEREDOC_STRINGS = ['$']
HEREDOC_TAG_IS_IDENTIFIER = True
HEREDOC_STRING_ALTERNATIVE = <TokenType.PARAMETER: 'PARAMETER'>
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.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'>, '~~~': <TokenType.GLOB: 'GLOB'>, '~~': <TokenType.LIKE: 'LIKE'>, '~~*': <TokenType.ILIKE: 'ILIKE'>, '~*': <TokenType.IRLIKE: 'IRLIKE'>, '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.BEGIN: 'BEGIN'>, 'BETWEEN': <TokenType.BETWEEN: 'BETWEEN'>, 'CACHE': <TokenType.CACHE: 'CACHE'>, 'UNCACHE': <TokenType.UNCACHE: 'UNCACHE'>, 'CASE': <TokenType.CASE: 'CASE'>, 'CHARACTER SET': <TokenType.CHARACTER_SET: 'CHARACTER_SET'>, 'CLUSTER BY': <TokenType.CLUSTER_BY: 'CLUSTER_BY'>, 'COLLATE': <TokenType.COLLATE: 'COLLATE'>, 'COLUMN': <TokenType.COLUMN: 'COLUMN'>, 'COMMIT': <TokenType.COMMIT: 'COMMIT'>, 'CONNECT BY': <TokenType.CONNECT_BY: 'CONNECT_BY'>, 'CONSTRAINT': <TokenType.CONSTRAINT: 'CONSTRAINT'>, 'COPY': <TokenType.COPY: 'COPY'>, 'CREATE': <TokenType.CREATE: 'CREATE'>, 'CROSS': <TokenType.CROSS: 'CROSS'>, 'CUBE': <TokenType.CUBE: 'CUBE'>, 'CURRENT_DATE': <TokenType.CURRENT_DATE: 'CURRENT_DATE'>, 'CURRENT_TIME': <TokenType.CURRENT_TIME: 'CURRENT_TIME'>, 'CURRENT_TIMESTAMP': <TokenType.CURRENT_TIMESTAMP: 'CURRENT_TIMESTAMP'>, 'CURRENT_USER': <TokenType.CURRENT_USER: 'CURRENT_USER'>, 'DATABASE': <TokenType.DATABASE: 'DATABASE'>, 'DEFAULT': <TokenType.DEFAULT: 'DEFAULT'>, 'DELETE': <TokenType.DELETE: 'DELETE'>, 'DESC': <TokenType.DESC: 'DESC'>, 'DESCRIBE': <TokenType.DESCRIBE: 'DESCRIBE'>, 'DISTINCT': <TokenType.DISTINCT: 'DISTINCT'>, 'DISTRIBUTE BY': <TokenType.DISTRIBUTE_BY: 'DISTRIBUTE_BY'>, 'DIV': <TokenType.DIV: 'DIV'>, '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'>, 'RENAME': <TokenType.RENAME: 'RENAME'>, 'REPLACE': <TokenType.REPLACE: 'REPLACE'>, 'RETURNING': <TokenType.RETURNING: 'RETURNING'>, 'REFERENCES': <TokenType.REFERENCES: 'REFERENCES'>, 'RIGHT': <TokenType.RIGHT: 'RIGHT'>, 'RLIKE': <TokenType.RLIKE: 'RLIKE'>, 'ROLLBACK': <TokenType.ROLLBACK: 'ROLLBACK'>, 'ROLLUP': <TokenType.ROLLUP: 'ROLLUP'>, 'ROW': <TokenType.ROW: 'ROW'>, 'ROWS': <TokenType.ROWS: 'ROWS'>, 'SCHEMA': <TokenType.SCHEMA: 'SCHEMA'>, 'SELECT': <TokenType.SELECT: 'SELECT'>, 'SEMI': <TokenType.SEMI: 'SEMI'>, 'SET': <TokenType.SET: 'SET'>, 'SETTINGS': <TokenType.SETTINGS: 'SETTINGS'>, 'SHOW': <TokenType.SHOW: 'SHOW'>, 'SIMILAR TO': <TokenType.SIMILAR_TO: 'SIMILAR_TO'>, 'SOME': <TokenType.SOME: 'SOME'>, 'SORT BY': <TokenType.SORT_BY: 'SORT_BY'>, 'START WITH': <TokenType.START_WITH: 'START_WITH'>, 'STRAIGHT_JOIN': <TokenType.STRAIGHT_JOIN: 'STRAIGHT_JOIN'>, 'TABLE': <TokenType.TABLE: 'TABLE'>, 'TABLESAMPLE': <TokenType.TABLE_SAMPLE: 'TABLE_SAMPLE'>, 'TEMP': <TokenType.TEMPORARY: 'TEMPORARY'>, 'TEMPORARY': <TokenType.TEMPORARY: 'TEMPORARY'>, 'THEN': <TokenType.THEN: 'THEN'>, 'TRUE': <TokenType.TRUE: 'TRUE'>, 'TRUNCATE': <TokenType.TRUNCATE: 'TRUNCATE'>, 'UNION': <TokenType.UNION: 'UNION'>, 'UNKNOWN': <TokenType.UNKNOWN: 'UNKNOWN'>, 'UNNEST': <TokenType.UNNEST: 'UNNEST'>, 'UNPIVOT': <TokenType.UNPIVOT: 'UNPIVOT'>, 'UPDATE': <TokenType.UPDATE: 'UPDATE'>, 'USE': <TokenType.USE: 'USE'>, 'USING': <TokenType.USING: 'USING'>, 'UUID': <TokenType.UUID: 'UUID'>, 'VALUES': <TokenType.VALUES: 'VALUES'>, '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'>, 'UHUGEINT': <TokenType.UINT128: 'UINT128'>, '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'>, 'DECIMAL32': <TokenType.DECIMAL32: 'DECIMAL32'>, 'DECIMAL64': <TokenType.DECIMAL64: 'DECIMAL64'>, 'DECIMAL128': <TokenType.DECIMAL128: 'DECIMAL128'>, 'DECIMAL256': <TokenType.DECIMAL256: 'DECIMAL256'>, 'BIGDECIMAL': <TokenType.BIGDECIMAL: 'BIGDECIMAL'>, 'BIGNUMERIC': <TokenType.BIGDECIMAL: 'BIGDECIMAL'>, 'LIST': <TokenType.LIST: 'LIST'>, 'MAP': <TokenType.MAP: 'MAP'>, 'NULLABLE': <TokenType.NULLABLE: 'NULLABLE'>, 'NUMBER': <TokenType.DECIMAL: 'DECIMAL'>, 'NUMERIC': <TokenType.DECIMAL: 'DECIMAL'>, 'FIXED': <TokenType.DECIMAL: 'DECIMAL'>, 'REAL': <TokenType.FLOAT: 'FLOAT'>, 'FLOAT': <TokenType.FLOAT: 'FLOAT'>, 'FLOAT4': <TokenType.FLOAT: 'FLOAT'>, 'FLOAT8': <TokenType.DOUBLE: 'DOUBLE'>, 'DOUBLE': <TokenType.DOUBLE: 'DOUBLE'>, 'DOUBLE PRECISION': <TokenType.DOUBLE: 'DOUBLE'>, 'JSON': <TokenType.JSON: 'JSON'>, 'JSONB': <TokenType.JSONB: 'JSONB'>, 'CHAR': <TokenType.TEXT: 'TEXT'>, 'CHARACTER': <TokenType.CHAR: 'CHAR'>, 'NCHAR': <TokenType.NCHAR: 'NCHAR'>, 'VARCHAR': <TokenType.TEXT: 'TEXT'>, 'VARCHAR2': <TokenType.VARCHAR: 'VARCHAR'>, 'NVARCHAR': <TokenType.NVARCHAR: 'NVARCHAR'>, 'NVARCHAR2': <TokenType.NVARCHAR: 'NVARCHAR'>, 'BPCHAR': <TokenType.TEXT: 'TEXT'>, '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.VARBINARY: 'VARBINARY'>, '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.TIMESTAMP: 'TIMESTAMP'>, 'TIMESTAMPTZ': <TokenType.TIMESTAMPTZ: 'TIMESTAMPTZ'>, 'TIMESTAMPLTZ': <TokenType.TIMESTAMPLTZ: 'TIMESTAMPLTZ'>, 'TIMESTAMP_LTZ': <TokenType.TIMESTAMPLTZ: 'TIMESTAMPLTZ'>, 'TIMESTAMPNTZ': <TokenType.TIMESTAMPNTZ: 'TIMESTAMPNTZ'>, 'TIMESTAMP_NTZ': <TokenType.TIMESTAMPNTZ: 'TIMESTAMPNTZ'>, 'DATE': <TokenType.DATE: 'DATE'>, 'DATETIME': <TokenType.DATETIME: 'DATETIME'>, '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'>, 'VECTOR': <TokenType.VECTOR: 'VECTOR'>, 'STRUCT': <TokenType.STRUCT: 'STRUCT'>, 'SEQUENCE': <TokenType.SEQUENCE: 'SEQUENCE'>, 'VARIANT': <TokenType.VARIANT: 'VARIANT'>, 'ALTER': <TokenType.ALTER: 'ALTER'>, 'ANALYZE': <TokenType.COMMAND: 'COMMAND'>, 'CALL': <TokenType.COMMAND: 'COMMAND'>, 'COMMENT': <TokenType.COMMENT: 'COMMENT'>, 'EXPLAIN': <TokenType.COMMAND: 'COMMAND'>, 'GRANT': <TokenType.GRANT: 'GRANT'>, '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'>, '//': <TokenType.DIV: 'DIV'>, '**': <TokenType.DSTAR: 'DSTAR'>, '^@': <TokenType.CARET_AT: 'CARET_AT'>, '@>': <TokenType.AT_GT: 'AT_GT'>, '<@': <TokenType.LT_AT: 'LT_AT'>, 'ATTACH': <TokenType.COMMAND: 'COMMAND'>, 'BITSTRING': <TokenType.BIT: 'BIT'>, 'CHARACTER VARYING': <TokenType.TEXT: 'TEXT'>, 'EXCLUDE': <TokenType.EXCEPT: 'EXCEPT'>, 'LOGICAL': <TokenType.BOOLEAN: 'BOOLEAN'>, 'ONLY': <TokenType.ONLY: 'ONLY'>, 'PIVOT_WIDER': <TokenType.PIVOT: 'PIVOT'>, 'POSITIONAL': <TokenType.POSITIONAL: 'POSITIONAL'>, 'SIGNED': <TokenType.INT: 'INT'>, 'SUMMARIZE': <TokenType.SUMMARIZE: 'SUMMARIZE'>, 'TIMESTAMP_S': <TokenType.TIMESTAMP_S: 'TIMESTAMP_S'>, 'TIMESTAMP_MS': <TokenType.TIMESTAMP_MS: 'TIMESTAMP_MS'>, 'TIMESTAMP_NS': <TokenType.TIMESTAMP_NS: 'TIMESTAMP_NS'>, 'TIMESTAMP_US': <TokenType.TIMESTAMP: 'TIMESTAMP'>, 'UBIGINT': <TokenType.UBIGINT: 'UBIGINT'>, 'UINTEGER': <TokenType.UINT: 'UINT'>, 'USMALLINT': <TokenType.USMALLINT: 'USMALLINT'>, 'UTINYINT': <TokenType.UTINYINT: 'UTINYINT'>}
SINGLE_TOKENS = {'(': <TokenType.L_PAREN: 'L_PAREN'>, ')': <TokenType.R_PAREN: 'R_PAREN'>, '[': <TokenType.L_BRACKET: 'L_BRACKET'>, ']': <TokenType.R_BRACKET: 'R_BRACKET'>, '{': <TokenType.L_BRACE: 'L_BRACE'>, '}': <TokenType.R_BRACE: 'R_BRACE'>, '&': <TokenType.AMP: 'AMP'>, '^': <TokenType.CARET: 'CARET'>, ':': <TokenType.COLON: 'COLON'>, ',': <TokenType.COMMA: 'COMMA'>, '.': <TokenType.DOT: 'DOT'>, '-': <TokenType.DASH: 'DASH'>, '=': <TokenType.EQ: 'EQ'>, '>': <TokenType.GT: 'GT'>, '<': <TokenType.LT: 'LT'>, '%': <TokenType.MOD: 'MOD'>, '!': <TokenType.NOT: 'NOT'>, '|': <TokenType.PIPE: 'PIPE'>, '+': <TokenType.PLUS: 'PLUS'>, ';': <TokenType.SEMICOLON: 'SEMICOLON'>, '/': <TokenType.SLASH: 'SLASH'>, '\\': <TokenType.BACKSLASH: 'BACKSLASH'>, '*': <TokenType.STAR: 'STAR'>, '~': <TokenType.TILDA: 'TILDA'>, '?': <TokenType.PLACEHOLDER: 'PLACEHOLDER'>, '@': <TokenType.PARAMETER: 'PARAMETER'>, '#': <TokenType.HASH: 'HASH'>, "'": <TokenType.UNKNOWN: 'UNKNOWN'>, '`': <TokenType.UNKNOWN: 'UNKNOWN'>, '"': <TokenType.UNKNOWN: 'UNKNOWN'>, '$': <TokenType.PARAMETER: 'PARAMETER'>}
class DuckDB.Parser(sqlglot.parser.Parser):
344    class Parser(parser.Parser):
345        BITWISE = {
346            **parser.Parser.BITWISE,
347            TokenType.TILDA: exp.RegexpLike,
348        }
349        BITWISE.pop(TokenType.CARET)
350
351        RANGE_PARSERS = {
352            **parser.Parser.RANGE_PARSERS,
353            TokenType.DAMP: binary_range_parser(exp.ArrayOverlaps),
354            TokenType.CARET_AT: binary_range_parser(exp.StartsWith),
355        }
356
357        EXPONENT = {
358            **parser.Parser.EXPONENT,
359            TokenType.CARET: exp.Pow,
360            TokenType.DSTAR: exp.Pow,
361        }
362
363        FUNCTIONS_WITH_ALIASED_ARGS = {*parser.Parser.FUNCTIONS_WITH_ALIASED_ARGS, "STRUCT_PACK"}
364
365        FUNCTIONS = {
366            **parser.Parser.FUNCTIONS,
367            "ARRAY_REVERSE_SORT": _build_sort_array_desc,
368            "ARRAY_SORT": exp.SortArray.from_arg_list,
369            "DATEDIFF": _build_date_diff,
370            "DATE_DIFF": _build_date_diff,
371            "DATE_TRUNC": date_trunc_to_time,
372            "DATETRUNC": date_trunc_to_time,
373            "DECODE": lambda args: exp.Decode(
374                this=seq_get(args, 0), charset=exp.Literal.string("utf-8")
375            ),
376            "ENCODE": lambda args: exp.Encode(
377                this=seq_get(args, 0), charset=exp.Literal.string("utf-8")
378            ),
379            "EPOCH": exp.TimeToUnix.from_arg_list,
380            "EPOCH_MS": lambda args: exp.UnixToTime(
381                this=seq_get(args, 0), scale=exp.UnixToTime.MILLIS
382            ),
383            "JSON": exp.ParseJSON.from_arg_list,
384            "JSON_EXTRACT_PATH": parser.build_extract_json_with_path(exp.JSONExtract),
385            "JSON_EXTRACT_STRING": parser.build_extract_json_with_path(exp.JSONExtractScalar),
386            "LIST_HAS": exp.ArrayContains.from_arg_list,
387            "LIST_REVERSE_SORT": _build_sort_array_desc,
388            "LIST_SORT": exp.SortArray.from_arg_list,
389            "LIST_VALUE": lambda args: exp.Array(expressions=args),
390            "MAKE_TIME": exp.TimeFromParts.from_arg_list,
391            "MAKE_TIMESTAMP": _build_make_timestamp,
392            "QUANTILE_CONT": exp.PercentileCont.from_arg_list,
393            "QUANTILE_DISC": exp.PercentileDisc.from_arg_list,
394            "REGEXP_EXTRACT": build_regexp_extract(exp.RegexpExtract),
395            "REGEXP_EXTRACT_ALL": build_regexp_extract(exp.RegexpExtractAll),
396            "REGEXP_MATCHES": exp.RegexpLike.from_arg_list,
397            "REGEXP_REPLACE": lambda args: exp.RegexpReplace(
398                this=seq_get(args, 0),
399                expression=seq_get(args, 1),
400                replacement=seq_get(args, 2),
401                modifiers=seq_get(args, 3),
402            ),
403            "STRFTIME": build_formatted_time(exp.TimeToStr, "duckdb"),
404            "STRING_SPLIT": exp.Split.from_arg_list,
405            "STRING_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
406            "STRING_TO_ARRAY": exp.Split.from_arg_list,
407            "STRPTIME": build_formatted_time(exp.StrToTime, "duckdb"),
408            "STRUCT_PACK": exp.Struct.from_arg_list,
409            "STR_SPLIT": exp.Split.from_arg_list,
410            "STR_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
411            "TO_TIMESTAMP": exp.UnixToTime.from_arg_list,
412            "UNNEST": exp.Explode.from_arg_list,
413            "XOR": binary_from_function(exp.BitwiseXor),
414            "GENERATE_SERIES": _build_generate_series(),
415            "RANGE": _build_generate_series(end_exclusive=True),
416            "EDITDIST3": exp.Levenshtein.from_arg_list,
417        }
418
419        FUNCTIONS.pop("DATE_SUB")
420        FUNCTIONS.pop("GLOB")
421
422        FUNCTION_PARSERS = parser.Parser.FUNCTION_PARSERS.copy()
423        FUNCTION_PARSERS.pop("DECODE")
424
425        NO_PAREN_FUNCTION_PARSERS = {
426            **parser.Parser.NO_PAREN_FUNCTION_PARSERS,
427            "MAP": lambda self: self._parse_map(),
428        }
429
430        TABLE_ALIAS_TOKENS = parser.Parser.TABLE_ALIAS_TOKENS - {
431            TokenType.SEMI,
432            TokenType.ANTI,
433        }
434
435        PLACEHOLDER_PARSERS = {
436            **parser.Parser.PLACEHOLDER_PARSERS,
437            TokenType.PARAMETER: lambda self: (
438                self.expression(exp.Placeholder, this=self._prev.text)
439                if self._match(TokenType.NUMBER) or self._match_set(self.ID_VAR_TOKENS)
440                else None
441            ),
442        }
443
444        TYPE_CONVERTERS = {
445            # https://duckdb.org/docs/sql/data_types/numeric
446            exp.DataType.Type.DECIMAL: build_default_decimal_type(precision=18, scale=3),
447            # https://duckdb.org/docs/sql/data_types/text
448            exp.DataType.Type.TEXT: lambda dtype: exp.DataType.build("TEXT"),
449        }
450
451        def _parse_table_sample(self, as_modifier: bool = False) -> t.Optional[exp.TableSample]:
452            # https://duckdb.org/docs/sql/samples.html
453            sample = super()._parse_table_sample(as_modifier=as_modifier)
454            if sample and not sample.args.get("method"):
455                if sample.args.get("size"):
456                    sample.set("method", exp.var("RESERVOIR"))
457                else:
458                    sample.set("method", exp.var("SYSTEM"))
459
460            return sample
461
462        def _parse_bracket(
463            self, this: t.Optional[exp.Expression] = None
464        ) -> t.Optional[exp.Expression]:
465            bracket = super()._parse_bracket(this)
466            if isinstance(bracket, exp.Bracket):
467                bracket.set("returns_list_for_maps", True)
468
469            return bracket
470
471        def _parse_map(self) -> exp.ToMap | exp.Map:
472            if self._match(TokenType.L_BRACE, advance=False):
473                return self.expression(exp.ToMap, this=self._parse_bracket())
474
475            args = self._parse_wrapped_csv(self._parse_assignment)
476            return self.expression(exp.Map, keys=seq_get(args, 0), values=seq_get(args, 1))
477
478        def _parse_struct_types(self, type_required: bool = False) -> t.Optional[exp.Expression]:
479            return self._parse_field_def()
480
481        def _pivot_column_names(self, aggregations: t.List[exp.Expression]) -> t.List[str]:
482            if len(aggregations) == 1:
483                return super()._pivot_column_names(aggregations)
484            return pivot_column_names(aggregations, dialect="duckdb")

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
BITWISE = {<TokenType.AMP: 'AMP'>: <class 'sqlglot.expressions.BitwiseAnd'>, <TokenType.PIPE: 'PIPE'>: <class 'sqlglot.expressions.BitwiseOr'>, <TokenType.TILDA: 'TILDA'>: <class 'sqlglot.expressions.RegexpLike'>}
RANGE_PARSERS = {<TokenType.AT_GT: 'AT_GT'>: <function binary_range_parser.<locals>._parse_binary_range>, <TokenType.BETWEEN: 'BETWEEN'>: <function Parser.<lambda>>, <TokenType.GLOB: 'GLOB'>: <function binary_range_parser.<locals>._parse_binary_range>, <TokenType.ILIKE: 'ILIKE'>: <function binary_range_parser.<locals>._parse_binary_range>, <TokenType.IN: 'IN'>: <function Parser.<lambda>>, <TokenType.IRLIKE: 'IRLIKE'>: <function binary_range_parser.<locals>._parse_binary_range>, <TokenType.IS: 'IS'>: <function Parser.<lambda>>, <TokenType.LIKE: 'LIKE'>: <function binary_range_parser.<locals>._parse_binary_range>, <TokenType.LT_AT: 'LT_AT'>: <function binary_range_parser.<locals>._parse_binary_range>, <TokenType.OVERLAPS: 'OVERLAPS'>: <function binary_range_parser.<locals>._parse_binary_range>, <TokenType.RLIKE: 'RLIKE'>: <function binary_range_parser.<locals>._parse_binary_range>, <TokenType.SIMILAR_TO: 'SIMILAR_TO'>: <function binary_range_parser.<locals>._parse_binary_range>, <TokenType.FOR: 'FOR'>: <function Parser.<lambda>>, <TokenType.DAMP: 'DAMP'>: <function binary_range_parser.<locals>._parse_binary_range>, <TokenType.CARET_AT: 'CARET_AT'>: <function binary_range_parser.<locals>._parse_binary_range>}
EXPONENT = {<TokenType.CARET: 'CARET'>: <class 'sqlglot.expressions.Pow'>, <TokenType.DSTAR: 'DSTAR'>: <class 'sqlglot.expressions.Pow'>}
FUNCTIONS_WITH_ALIASED_ARGS = {'STRUCT', 'STRUCT_PACK'}
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'>>, 'APPLY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Apply'>>, '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': <function Parser.<lambda>>, 'ARRAY_AGG': <function Parser.<lambda>>, 'ARRAY_ALL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayAll'>>, 'ARRAY_ANY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayAny'>>, 'ARRAY_CONCAT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayConcat'>>, 'ARRAY_CAT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayConcat'>>, 'ARRAY_CONSTRUCT_COMPACT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayConstructCompact'>>, 'ARRAY_CONTAINS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayContains'>>, 'ARRAY_HAS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayContains'>>, 'ARRAY_CONTAINS_ALL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayContainsAll'>>, 'ARRAY_HAS_ALL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayContainsAll'>>, 'FILTER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayFilter'>>, 'ARRAY_FILTER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayFilter'>>, 'ARRAY_OVERLAPS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayOverlaps'>>, 'ARRAY_SIZE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArraySize'>>, 'ARRAY_LENGTH': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArraySize'>>, 'ARRAY_SORT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.SortArray'>>, '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': <function Parser.<lambda>>, 'CHAR': <function Parser.<lambda>>, 'COALESCE': <function build_coalesce>, 'IFNULL': <function build_coalesce>, 'NVL': <function build_coalesce>, 'COLLATE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Collate'>>, 'COLUMNS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Columns'>>, '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'>>, 'CONVERT_TIMEZONE': <function build_convert_timezone>, 'CORR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Corr'>>, 'COUNT': <function Parser.<lambda>>, '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': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Date'>>, 'DATE_ADD': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DateAdd'>>, 'DATEDIFF': <function _build_date_diff>, 'DATE_DIFF': <function _build_date_diff>, '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_TO_DATE_STR': <function Parser.<lambda>>, 'DATE_TO_DI': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DateToDi'>>, 'DATE_TRUNC': <function date_trunc_to_time>, 'DATETIME': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Datetime'>>, 'DATETIME_ADD': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DatetimeAdd'>>, 'DATETIME_DIFF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DatetimeDiff'>>, 'DATETIME_SUB': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DatetimeSub'>>, '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'>>, 'DAYOFWEEK_ISO': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DayOfWeekIso'>>, 'ISODOW': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DayOfWeekIso'>>, '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': <function DuckDB.Parser.<lambda>>, 'DI_TO_DATE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DiToDate'>>, 'ENCODE': <function DuckDB.Parser.<lambda>>, 'EXISTS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Exists'>>, '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'>>, 'EXPLODING_GENERATE_SERIES': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ExplodingGenerateSeries'>>, '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'>>, 'FROM_ISO8601_TIMESTAMP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.FromISO8601Timestamp'>>, 'GAP_FILL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.GapFill'>>, 'GENERATE_DATE_ARRAY': <function Parser.<lambda>>, 'GENERATE_SERIES': <function _build_generate_series.<locals>._builder>, 'GENERATE_TIMESTAMP_ARRAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.GenerateTimestampArray'>>, 'GREATEST': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Greatest'>>, 'GROUP_CONCAT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.GroupConcat'>>, 'HEX': <function build_hex>, 'HLL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Hll'>>, 'IF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.If'>>, 'IIF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.If'>>, 'INITCAP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Initcap'>>, 'INLINE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Inline'>>, 'IS_INF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.IsInf'>>, 'ISINF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.IsInf'>>, 'IS_NAN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.IsNan'>>, 'ISNAN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.IsNan'>>, 'J_S_O_N_ARRAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONArray'>>, 'J_S_O_N_ARRAY_AGG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONArrayAgg'>>, 'JSON_ARRAY_CONTAINS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONArrayContains'>>, 'JSONB_CONTAINS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONBContains'>>, 'JSONB_EXISTS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONBExists'>>, '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'>>, 'J_S_O_N_EXISTS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONExists'>>, 'JSON_EXTRACT': <function build_extract_json_with_path.<locals>._builder>, 'J_S_O_N_EXTRACT_ARRAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONExtractArray'>>, 'JSON_EXTRACT_SCALAR': <function build_extract_json_with_path.<locals>._builder>, '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'>>, 'J_S_O_N_VALUE_ARRAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONValueArray'>>, 'LAG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Lag'>>, 'LAST': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Last'>>, 'LAST_DAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LastDay'>>, 'LAST_DAY_OF_MONTH': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LastDay'>>, 'LAST_VALUE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LastValue'>>, 'LEAD': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Lead'>>, 'LEAST': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Least'>>, 'LEFT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Left'>>, 'LENGTH': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Length'>>, 'LEN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Length'>>, 'LEVENSHTEIN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Levenshtein'>>, 'LIST': <bound method Func.from_arg_list of <class 'sqlglot.expressions.List'>>, 'LN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Ln'>>, 'LOG': <function build_logarithm>, 'LOGICAL_AND': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LogicalAnd'>>, 'BOOL_AND': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LogicalAnd'>>, 'BOOLAND_AGG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LogicalAnd'>>, 'LOGICAL_OR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LogicalOr'>>, 'BOOL_OR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LogicalOr'>>, 'BOOLOR_AGG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LogicalOr'>>, 'LOWER': <function build_lower>, 'LCASE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Lower'>>, 'LOWER_HEX': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LowerHex'>>, 'MD5': <bound method Func.from_arg_list of <class 'sqlglot.expressions.MD5'>>, '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'>>, 'MEDIAN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Median'>>, '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'>>, 'NORMALIZE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Normalize'>>, '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'>>, 'OBJECT_INSERT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ObjectInsert'>>, 'OPEN_J_S_O_N': <bound method Func.from_arg_list of <class 'sqlglot.expressions.OpenJSON'>>, 'OVERLAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Overlay'>>, 'PAD': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Pad'>>, 'PARAMETERIZED_AGG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ParameterizedAgg'>>, 'PARSE_JSON': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ParseJSON'>>, 'JSON_PARSE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ParseJSON'>>, 'PERCENTILE_CONT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.PercentileCont'>>, 'PERCENTILE_DISC': <bound method Func.from_arg_list of <class 'sqlglot.expressions.PercentileDisc'>>, 'POSEXPLODE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Posexplode'>>, 'POSEXPLODE_OUTER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.PosexplodeOuter'>>, 'POWER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Pow'>>, 'POW': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Pow'>>, 'PREDICT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Predict'>>, 'QUANTILE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Quantile'>>, 'QUARTER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Quarter'>>, 'RAND': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Rand'>>, 'RANDOM': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Rand'>>, 'RANDN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Randn'>>, 'RANGE_N': <bound method Func.from_arg_list of <class 'sqlglot.expressions.RangeN'>>, 'READ_CSV': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ReadCSV'>>, 'REDUCE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Reduce'>>, 'REGEXP_EXTRACT': <function build_regexp_extract.<locals>._builder>, 'REGEXP_EXTRACT_ALL': <function build_regexp_extract.<locals>._builder>, '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': <function DuckDB.Parser.<lambda>>, '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': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Split'>>, 'SPLIT_PART': <bound method Func.from_arg_list of <class 'sqlglot.expressions.SplitPart'>>, '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'>>, 'STDEV': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Stddev'>>, 'STDDEV_POP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StddevPop'>>, 'STDDEV_SAMP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StddevSamp'>>, 'STR_POSITION': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StrPosition'>>, 'STR_TO_DATE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StrToDate'>>, 'STR_TO_MAP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StrToMap'>>, 'STR_TO_TIME': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StrToTime'>>, 'STR_TO_UNIX': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StrToUnix'>>, 'STRING': <bound method Func.from_arg_list of <class 'sqlglot.expressions.String'>>, 'STRING_TO_ARRAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Split'>>, 'SPLIT_BY_STRING': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StringToArray'>>, 'STRUCT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Struct'>>, 'STRUCT_EXTRACT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StructExtract'>>, 'STUFF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Stuff'>>, 'INSERT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Stuff'>>, 'SUBSTRING': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Substring'>>, 'SUBSTR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Substring'>>, 'SUM': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Sum'>>, 'TIME': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Time'>>, 'TIME_ADD': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TimeAdd'>>, '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': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TimeSub'>>, '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': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Timestamp'>>, 'TIMESTAMP_ADD': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TimestampAdd'>>, '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': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TimestampSub'>>, '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_DOUBLE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ToDouble'>>, 'TO_MAP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ToMap'>>, 'TO_NUMBER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ToNumber'>>, 'TRANSFORM': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Transform'>>, 'TRIM': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Trim'>>, 'TRY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Try'>>, 'TRY_CAST': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TryCast'>>, 'TS_OR_DI_TO_DI': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDiToDi'>>, 'TS_OR_DS_ADD': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDsAdd'>>, 'TS_OR_DS_DIFF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDsDiff'>>, 'TS_OR_DS_TO_DATE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDsToDate'>>, 'TS_OR_DS_TO_DATE_STR': <function Parser.<lambda>>, 'TS_OR_DS_TO_TIME': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDsToTime'>>, 'TS_OR_DS_TO_TIMESTAMP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDsToTimestamp'>>, 'UNHEX': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Unhex'>>, 'UNIX_DATE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.UnixDate'>>, 'UNIX_SECONDS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.UnixSeconds'>>, 'UNIX_TO_STR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.UnixToStr'>>, 'UNIX_TO_TIME': <bound method Func.from_arg_list of <class 'sqlglot.expressions.UnixToTime'>>, 'UNIX_TO_TIME_STR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.UnixToTimeStr'>>, 'UNNEST': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Explode'>>, 'UPPER': <function build_upper>, 'UCASE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Upper'>>, 'UUID': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Uuid'>>, 'GEN_RANDOM_UUID': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Uuid'>>, 'GENERATE_UUID': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Uuid'>>, 'UUID_STRING': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Uuid'>>, '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': <function binary_from_function.<locals>.<lambda>>, 'YEAR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Year'>>, 'ARRAYAGG': <function Parser.<lambda>>, 'INSTR': <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>>, 'LPAD': <function Parser.<lambda>>, 'LEFTPAD': <function Parser.<lambda>>, 'LTRIM': <function Parser.<lambda>>, 'MOD': <function build_mod>, 'RIGHTPAD': <function Parser.<lambda>>, 'RPAD': <function Parser.<lambda>>, 'RTRIM': <function Parser.<lambda>>, 'SCOPE_RESOLUTION': <function Parser.<lambda>>, 'TO_HEX': <function build_hex>, 'ARRAY_REVERSE_SORT': <function _build_sort_array_desc>, 'DATETRUNC': <function date_trunc_to_time>, 'EPOCH': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TimeToUnix'>>, 'EPOCH_MS': <function DuckDB.Parser.<lambda>>, 'JSON': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ParseJSON'>>, 'JSON_EXTRACT_PATH': <function build_extract_json_with_path.<locals>._builder>, 'JSON_EXTRACT_STRING': <function build_extract_json_with_path.<locals>._builder>, 'LIST_HAS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayContains'>>, 'LIST_REVERSE_SORT': <function _build_sort_array_desc>, 'LIST_SORT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.SortArray'>>, 'LIST_VALUE': <function DuckDB.Parser.<lambda>>, 'MAKE_TIME': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TimeFromParts'>>, 'MAKE_TIMESTAMP': <function _build_make_timestamp>, 'QUANTILE_CONT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.PercentileCont'>>, 'QUANTILE_DISC': <bound method Func.from_arg_list of <class 'sqlglot.expressions.PercentileDisc'>>, 'REGEXP_MATCHES': <bound method Func.from_arg_list of <class 'sqlglot.expressions.RegexpLike'>>, 'STRFTIME': <function build_formatted_time.<locals>._builder>, 'STRING_SPLIT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Split'>>, 'STRING_SPLIT_REGEX': <bound method Func.from_arg_list of <class 'sqlglot.expressions.RegexpSplit'>>, 'STRPTIME': <function build_formatted_time.<locals>._builder>, 'STRUCT_PACK': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Struct'>>, 'STR_SPLIT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Split'>>, 'STR_SPLIT_REGEX': <bound method Func.from_arg_list of <class 'sqlglot.expressions.RegexpSplit'>>, 'TO_TIMESTAMP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.UnixToTime'>>, 'RANGE': <function _build_generate_series.<locals>._builder>, 'EDITDIST3': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Levenshtein'>>}
FUNCTION_PARSERS = {'CAST': <function Parser.<lambda>>, 'CONVERT': <function Parser.<lambda>>, 'EXTRACT': <function Parser.<lambda>>, 'GAP_FILL': <function Parser.<lambda>>, 'JSON_OBJECT': <function Parser.<lambda>>, 'JSON_OBJECTAGG': <function Parser.<lambda>>, 'JSON_TABLE': <function Parser.<lambda>>, 'MATCH': <function Parser.<lambda>>, 'NORMALIZE': <function Parser.<lambda>>, 'OPENJSON': <function Parser.<lambda>>, 'OVERLAY': <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>>, 'TRIM': <function Parser.<lambda>>, 'TRY_CAST': <function Parser.<lambda>>, 'TRY_CONVERT': <function Parser.<lambda>>}
NO_PAREN_FUNCTION_PARSERS = {'ANY': <function Parser.<lambda>>, 'CASE': <function Parser.<lambda>>, 'CONNECT_BY_ROOT': <function Parser.<lambda>>, 'IF': <function Parser.<lambda>>, 'NEXT': <function Parser.<lambda>>, 'MAP': <function DuckDB.Parser.<lambda>>}
TABLE_ALIAS_TOKENS = {<TokenType.BOOLEAN: 'BOOLEAN'>, <TokenType.VOLATILE: 'VOLATILE'>, <TokenType.DECIMAL64: 'DECIMAL64'>, <TokenType.UNIQUEIDENTIFIER: 'UNIQUEIDENTIFIER'>, <TokenType.PSEUDO_TYPE: 'PSEUDO_TYPE'>, <TokenType.MODEL: 'MODEL'>, <TokenType.TSMULTIRANGE: 'TSMULTIRANGE'>, <TokenType.STREAMLIT: 'STREAMLIT'>, <TokenType.UINT: 'UINT'>, <TokenType.FIRST: 'FIRST'>, <TokenType.AUTO_INCREMENT: 'AUTO_INCREMENT'>, <TokenType.LIST: 'LIST'>, <TokenType.HLLSKETCH: 'HLLSKETCH'>, <TokenType.OBJECT: 'OBJECT'>, <TokenType.UUID: 'UUID'>, <TokenType.SOME: 'SOME'>, <TokenType.TSRANGE: 'TSRANGE'>, <TokenType.TABLE: 'TABLE'>, <TokenType.UNKNOWN: 'UNKNOWN'>, <TokenType.UMEDIUMINT: 'UMEDIUMINT'>, <TokenType.TSTZRANGE: 'TSTZRANGE'>, <TokenType.ORDINALITY: 'ORDINALITY'>, <TokenType.SMALLSERIAL: 'SMALLSERIAL'>, <TokenType.DICTIONARY: 'DICTIONARY'>, <TokenType.NULLABLE: 'NULLABLE'>, <TokenType.UNIQUE: 'UNIQUE'>, <TokenType.REFRESH: 'REFRESH'>, <TokenType.TIMESTAMP_S: 'TIMESTAMP_S'>, <TokenType.GEOGRAPHY: 'GEOGRAPHY'>, <TokenType.FOREIGN_KEY: 'FOREIGN_KEY'>, <TokenType.ENUM8: 'ENUM8'>, <TokenType.TIMESTAMPNTZ: 'TIMESTAMPNTZ'>, <TokenType.AGGREGATEFUNCTION: 'AGGREGATEFUNCTION'>, <TokenType.PARTITION: 'PARTITION'>, <TokenType.LONGBLOB: 'LONGBLOB'>, <TokenType.INT8RANGE: 'INT8RANGE'>, <TokenType.SMALLINT: 'SMALLINT'>, <TokenType.INDEX: 'INDEX'>, <TokenType.NAME: 'NAME'>, <TokenType.SHOW: 'SHOW'>, <TokenType.STRUCT: 'STRUCT'>, <TokenType.FINAL: 'FINAL'>, <TokenType.BIGDECIMAL: 'BIGDECIMAL'>, <TokenType.ENUM: 'ENUM'>, <TokenType.LOAD: 'LOAD'>, <TokenType.DATETIME: 'DATETIME'>, <TokenType.TIMESTAMPTZ: 'TIMESTAMPTZ'>, <TokenType.UPDATE: 'UPDATE'>, <TokenType.PERCENT: 'PERCENT'>, <TokenType.SMALLMONEY: 'SMALLMONEY'>, <TokenType.FALSE: 'FALSE'>, <TokenType.BIT: 'BIT'>, <TokenType.IPADDRESS: 'IPADDRESS'>, <TokenType.UINT128: 'UINT128'>, <TokenType.INT4RANGE: 'INT4RANGE'>, <TokenType.CONSTRAINT: 'CONSTRAINT'>, <TokenType.SET: 'SET'>, <TokenType.COMMIT: 'COMMIT'>, <TokenType.TEMPORARY: 'TEMPORARY'>, <TokenType.EXISTS: 'EXISTS'>, <TokenType.DATE: 'DATE'>, <TokenType.USMALLINT: 'USMALLINT'>, <TokenType.DATERANGE: 'DATERANGE'>, <TokenType.FORMAT: 'FORMAT'>, <TokenType.NUMMULTIRANGE: 'NUMMULTIRANGE'>, <TokenType.DATEMULTIRANGE: 'DATEMULTIRANGE'>, <TokenType.GEOMETRY: 'GEOMETRY'>, <TokenType.TINYTEXT: 'TINYTEXT'>, <TokenType.VARCHAR: 'VARCHAR'>, <TokenType.CURRENT_TIMESTAMP: 'CURRENT_TIMESTAMP'>, <TokenType.INT128: 'INT128'>, <TokenType.RECURSIVE: 'RECURSIVE'>, <TokenType.STRAIGHT_JOIN: 'STRAIGHT_JOIN'>, <TokenType.ARRAY: 'ARRAY'>, <TokenType.SERIAL: 'SERIAL'>, <TokenType.VIEW: 'VIEW'>, <TokenType.JSONB: 'JSONB'>, <TokenType.TIMESTAMP: 'TIMESTAMP'>, <TokenType.HSTORE: 'HSTORE'>, <TokenType.CURRENT_USER: 'CURRENT_USER'>, <TokenType.INT: 'INT'>, <TokenType.CURRENT_DATE: 'CURRENT_DATE'>, <TokenType.POINT: 'POINT'>, <TokenType.DECIMAL: 'DECIMAL'>, <TokenType.DEFAULT: 'DEFAULT'>, <TokenType.SCHEMA: 'SCHEMA'>, <TokenType.ISNULL: 'ISNULL'>, <TokenType.BEGIN: 'BEGIN'>, <TokenType.INT4MULTIRANGE: 'INT4MULTIRANGE'>, <TokenType.ROWS: 'ROWS'>, <TokenType.IMAGE: 'IMAGE'>, <TokenType.INTERVAL: 'INTERVAL'>, <TokenType.MULTILINESTRING: 'MULTILINESTRING'>, <TokenType.VARIANT: 'VARIANT'>, <TokenType.PIVOT: 'PIVOT'>, <TokenType.DOUBLE: 'DOUBLE'>, <TokenType.BPCHAR: 'BPCHAR'>, <TokenType.TRUNCATE: 'TRUNCATE'>, <TokenType.OBJECT_IDENTIFIER: 'OBJECT_IDENTIFIER'>, <TokenType.NESTED: 'NESTED'>, <TokenType.BINARY: 'BINARY'>, <TokenType.RENAME: 'RENAME'>, <TokenType.INT256: 'INT256'>, <TokenType.TIMESTAMP_MS: 'TIMESTAMP_MS'>, <TokenType.REFERENCES: 'REFERENCES'>, <TokenType.UTINYINT: 'UTINYINT'>, <TokenType.DATE32: 'DATE32'>, <TokenType.JSON: 'JSON'>, <TokenType.DIV: 'DIV'>, <TokenType.KEEP: 'KEEP'>, <TokenType.OVERWRITE: 'OVERWRITE'>, <TokenType.TOP: 'TOP'>, <TokenType.MEDIUMINT: 'MEDIUMINT'>, <TokenType.NVARCHAR: 'NVARCHAR'>, <TokenType.COLLATE: 'COLLATE'>, <TokenType.DESCRIBE: 'DESCRIBE'>, <TokenType.ROWVERSION: 'ROWVERSION'>, <TokenType.CURRENT_TIME: 'CURRENT_TIME'>, <TokenType.IPPREFIX: 'IPPREFIX'>, <TokenType.DESC: 'DESC'>, <TokenType.TIMESTAMP_NS: 'TIMESTAMP_NS'>, <TokenType.TRUE: 'TRUE'>, <TokenType.SUPER: 'SUPER'>, <TokenType.MAP: 'MAP'>, <TokenType.NEXT: 'NEXT'>, <TokenType.REPLACE: 'REPLACE'>, <TokenType.COLUMN: 'COLUMN'>, <TokenType.UNPIVOT: 'UNPIVOT'>, <TokenType.TEXT: 'TEXT'>, <TokenType.POLYGON: 'POLYGON'>, <TokenType.USERDEFINED: 'USERDEFINED'>, <TokenType.MONEY: 'MONEY'>, <TokenType.ROW: 'ROW'>, <TokenType.TSTZMULTIRANGE: 'TSTZMULTIRANGE'>, <TokenType.VARBINARY: 'VARBINARY'>, <TokenType.OPERATOR: 'OPERATOR'>, <TokenType.PROCEDURE: 'PROCEDURE'>, <TokenType.DATABASE: 'DATABASE'>, <TokenType.TIMESTAMPLTZ: 'TIMESTAMPLTZ'>, <TokenType.ALL: 'ALL'>, <TokenType.UDECIMAL: 'UDECIMAL'>, <TokenType.WAREHOUSE: 'WAREHOUSE'>, <TokenType.EXECUTE: 'EXECUTE'>, <TokenType.XML: 'XML'>, <TokenType.DECIMAL32: 'DECIMAL32'>, <TokenType.FLOAT: 'FLOAT'>, <TokenType.COPY: 'COPY'>, <TokenType.TINYINT: 'TINYINT'>, <TokenType.KILL: 'KILL'>, <TokenType.TINYBLOB: 'TINYBLOB'>, <TokenType.MERGE: 'MERGE'>, <TokenType.OVERLAPS: 'OVERLAPS'>, <TokenType.DECIMAL128: 'DECIMAL128'>, <TokenType.BIGSERIAL: 'BIGSERIAL'>, <TokenType.UBIGINT: 'UBIGINT'>, <TokenType.ENUM16: 'ENUM16'>, <TokenType.CACHE: 'CACHE'>, <TokenType.NUMRANGE: 'NUMRANGE'>, <TokenType.BIGINT: 'BIGINT'>, <TokenType.SEQUENCE: 'SEQUENCE'>, <TokenType.LOWCARDINALITY: 'LOWCARDINALITY'>, <TokenType.IS: 'IS'>, <TokenType.NCHAR: 'NCHAR'>, <TokenType.ROLLUP: 'ROLLUP'>, <TokenType.DECIMAL256: 'DECIMAL256'>, <TokenType.FUNCTION: 'FUNCTION'>, <TokenType.LINESTRING: 'LINESTRING'>, <TokenType.IPV4: 'IPV4'>, <TokenType.STORAGE_INTEGRATION: 'STORAGE_INTEGRATION'>, <TokenType.INET: 'INET'>, <TokenType.RING: 'RING'>, <TokenType.TIME: 'TIME'>, <TokenType.RANGE: 'RANGE'>, <TokenType.FIXEDSTRING: 'FIXEDSTRING'>, <TokenType.FILTER: 'FILTER'>, <TokenType.VAR: 'VAR'>, <TokenType.IPV6: 'IPV6'>, <TokenType.TAG: 'TAG'>, <TokenType.UINT256: 'UINT256'>, <TokenType.CASE: 'CASE'>, <TokenType.TDIGEST: 'TDIGEST'>, <TokenType.SETTINGS: 'SETTINGS'>, <TokenType.YEAR: 'YEAR'>, <TokenType.ANY: 'ANY'>, <TokenType.CURRENT_DATETIME: 'CURRENT_DATETIME'>, <TokenType.USE: 'USE'>, <TokenType.LONGTEXT: 'LONGTEXT'>, <TokenType.DATETIME64: 'DATETIME64'>, <TokenType.COMMENT: 'COMMENT'>, <TokenType.DELETE: 'DELETE'>, <TokenType.ESCAPE: 'ESCAPE'>, <TokenType.END: 'END'>, <TokenType.CHAR: 'CHAR'>, <TokenType.IDENTIFIER: 'IDENTIFIER'>, <TokenType.MULTIPOLYGON: 'MULTIPOLYGON'>, <TokenType.COMMAND: 'COMMAND'>, <TokenType.CUBE: 'CUBE'>, <TokenType.SIMPLEAGGREGATEFUNCTION: 'SIMPLEAGGREGATEFUNCTION'>, <TokenType.INT8MULTIRANGE: 'INT8MULTIRANGE'>, <TokenType.PRAGMA: 'PRAGMA'>, <TokenType.ASC: 'ASC'>, <TokenType.NULL: 'NULL'>, <TokenType.TIMETZ: 'TIMETZ'>, <TokenType.VECTOR: 'VECTOR'>, <TokenType.MEDIUMBLOB: 'MEDIUMBLOB'>, <TokenType.MEDIUMTEXT: 'MEDIUMTEXT'>, <TokenType.UNNEST: 'UNNEST'>}
PLACEHOLDER_PARSERS = {<TokenType.PLACEHOLDER: 'PLACEHOLDER'>: <function Parser.<lambda>>, <TokenType.PARAMETER: 'PARAMETER'>: <function DuckDB.Parser.<lambda>>, <TokenType.COLON: 'COLON'>: <function Parser.<lambda>>}
TYPE_CONVERTERS = {<Type.DECIMAL: 'DECIMAL'>: <function build_default_decimal_type.<locals>._builder>, <Type.TEXT: 'TEXT'>: <function DuckDB.Parser.<lambda>>}
ID_VAR_TOKENS = {<TokenType.BOOLEAN: 'BOOLEAN'>, <TokenType.ASOF: 'ASOF'>, <TokenType.VOLATILE: 'VOLATILE'>, <TokenType.DECIMAL64: 'DECIMAL64'>, <TokenType.UNIQUEIDENTIFIER: 'UNIQUEIDENTIFIER'>, <TokenType.PSEUDO_TYPE: 'PSEUDO_TYPE'>, <TokenType.MODEL: 'MODEL'>, <TokenType.TSMULTIRANGE: 'TSMULTIRANGE'>, <TokenType.WINDOW: 'WINDOW'>, <TokenType.STREAMLIT: 'STREAMLIT'>, <TokenType.UINT: 'UINT'>, <TokenType.FIRST: 'FIRST'>, <TokenType.AUTO_INCREMENT: 'AUTO_INCREMENT'>, <TokenType.LIST: 'LIST'>, <TokenType.HLLSKETCH: 'HLLSKETCH'>, <TokenType.OBJECT: 'OBJECT'>, <TokenType.UUID: 'UUID'>, <TokenType.SOME: 'SOME'>, <TokenType.TSRANGE: 'TSRANGE'>, <TokenType.TABLE: 'TABLE'>, <TokenType.UNKNOWN: 'UNKNOWN'>, <TokenType.UMEDIUMINT: 'UMEDIUMINT'>, <TokenType.TSTZRANGE: 'TSTZRANGE'>, <TokenType.ORDINALITY: 'ORDINALITY'>, <TokenType.SMALLSERIAL: 'SMALLSERIAL'>, <TokenType.DICTIONARY: 'DICTIONARY'>, <TokenType.NULLABLE: 'NULLABLE'>, <TokenType.UNIQUE: 'UNIQUE'>, <TokenType.REFRESH: 'REFRESH'>, <TokenType.ANTI: 'ANTI'>, <TokenType.TIMESTAMP_S: 'TIMESTAMP_S'>, <TokenType.GEOGRAPHY: 'GEOGRAPHY'>, <TokenType.FOREIGN_KEY: 'FOREIGN_KEY'>, <TokenType.ENUM8: 'ENUM8'>, <TokenType.OFFSET: 'OFFSET'>, <TokenType.TIMESTAMPNTZ: 'TIMESTAMPNTZ'>, <TokenType.AGGREGATEFUNCTION: 'AGGREGATEFUNCTION'>, <TokenType.PARTITION: 'PARTITION'>, <TokenType.LONGBLOB: 'LONGBLOB'>, <TokenType.INT8RANGE: 'INT8RANGE'>, <TokenType.SMALLINT: 'SMALLINT'>, <TokenType.INDEX: 'INDEX'>, <TokenType.NAME: 'NAME'>, <TokenType.SHOW: 'SHOW'>, <TokenType.STRUCT: 'STRUCT'>, <TokenType.FINAL: 'FINAL'>, <TokenType.BIGDECIMAL: 'BIGDECIMAL'>, <TokenType.ENUM: 'ENUM'>, <TokenType.FULL: 'FULL'>, <TokenType.LOAD: 'LOAD'>, <TokenType.DATETIME: 'DATETIME'>, <TokenType.TIMESTAMPTZ: 'TIMESTAMPTZ'>, <TokenType.UPDATE: 'UPDATE'>, <TokenType.PERCENT: 'PERCENT'>, <TokenType.SMALLMONEY: 'SMALLMONEY'>, <TokenType.FALSE: 'FALSE'>, <TokenType.BIT: 'BIT'>, <TokenType.IPADDRESS: 'IPADDRESS'>, <TokenType.UINT128: 'UINT128'>, <TokenType.INT4RANGE: 'INT4RANGE'>, <TokenType.CONSTRAINT: 'CONSTRAINT'>, <TokenType.SET: 'SET'>, <TokenType.COMMIT: 'COMMIT'>, <TokenType.TEMPORARY: 'TEMPORARY'>, <TokenType.EXISTS: 'EXISTS'>, <TokenType.DATE: 'DATE'>, <TokenType.USMALLINT: 'USMALLINT'>, <TokenType.DATERANGE: 'DATERANGE'>, <TokenType.FORMAT: 'FORMAT'>, <TokenType.NUMMULTIRANGE: 'NUMMULTIRANGE'>, <TokenType.DATEMULTIRANGE: 'DATEMULTIRANGE'>, <TokenType.GEOMETRY: 'GEOMETRY'>, <TokenType.LEFT: 'LEFT'>, <TokenType.TINYTEXT: 'TINYTEXT'>, <TokenType.VARCHAR: 'VARCHAR'>, <TokenType.CURRENT_TIMESTAMP: 'CURRENT_TIMESTAMP'>, <TokenType.APPLY: 'APPLY'>, <TokenType.RECURSIVE: 'RECURSIVE'>, <TokenType.INT128: 'INT128'>, <TokenType.STRAIGHT_JOIN: 'STRAIGHT_JOIN'>, <TokenType.ARRAY: 'ARRAY'>, <TokenType.SERIAL: 'SERIAL'>, <TokenType.VIEW: 'VIEW'>, <TokenType.JSONB: 'JSONB'>, <TokenType.TIMESTAMP: 'TIMESTAMP'>, <TokenType.HSTORE: 'HSTORE'>, <TokenType.CURRENT_USER: 'CURRENT_USER'>, <TokenType.INT: 'INT'>, <TokenType.CURRENT_DATE: 'CURRENT_DATE'>, <TokenType.POINT: 'POINT'>, <TokenType.DECIMAL: 'DECIMAL'>, <TokenType.DEFAULT: 'DEFAULT'>, <TokenType.SCHEMA: 'SCHEMA'>, <TokenType.ISNULL: 'ISNULL'>, <TokenType.BEGIN: 'BEGIN'>, <TokenType.INT4MULTIRANGE: 'INT4MULTIRANGE'>, <TokenType.ROWS: 'ROWS'>, <TokenType.IMAGE: 'IMAGE'>, <TokenType.INTERVAL: 'INTERVAL'>, <TokenType.MULTILINESTRING: 'MULTILINESTRING'>, <TokenType.VARIANT: 'VARIANT'>, <TokenType.PIVOT: 'PIVOT'>, <TokenType.DOUBLE: 'DOUBLE'>, <TokenType.BPCHAR: 'BPCHAR'>, <TokenType.TRUNCATE: 'TRUNCATE'>, <TokenType.OBJECT_IDENTIFIER: 'OBJECT_IDENTIFIER'>, <TokenType.NESTED: 'NESTED'>, <TokenType.BINARY: 'BINARY'>, <TokenType.NATURAL: 'NATURAL'>, <TokenType.RENAME: 'RENAME'>, <TokenType.INT256: 'INT256'>, <TokenType.RIGHT: 'RIGHT'>, <TokenType.TIMESTAMP_MS: 'TIMESTAMP_MS'>, <TokenType.REFERENCES: 'REFERENCES'>, <TokenType.UTINYINT: 'UTINYINT'>, <TokenType.DATE32: 'DATE32'>, <TokenType.JSON: 'JSON'>, <TokenType.DIV: 'DIV'>, <TokenType.KEEP: 'KEEP'>, <TokenType.OVERWRITE: 'OVERWRITE'>, <TokenType.TOP: 'TOP'>, <TokenType.MEDIUMINT: 'MEDIUMINT'>, <TokenType.SEMI: 'SEMI'>, <TokenType.NVARCHAR: 'NVARCHAR'>, <TokenType.COLLATE: 'COLLATE'>, <TokenType.DESCRIBE: 'DESCRIBE'>, <TokenType.ROWVERSION: 'ROWVERSION'>, <TokenType.CURRENT_TIME: 'CURRENT_TIME'>, <TokenType.IPPREFIX: 'IPPREFIX'>, <TokenType.DESC: 'DESC'>, <TokenType.TIMESTAMP_NS: 'TIMESTAMP_NS'>, <TokenType.TRUE: 'TRUE'>, <TokenType.SUPER: 'SUPER'>, <TokenType.MAP: 'MAP'>, <TokenType.NEXT: 'NEXT'>, <TokenType.REPLACE: 'REPLACE'>, <TokenType.COLUMN: 'COLUMN'>, <TokenType.UNPIVOT: 'UNPIVOT'>, <TokenType.TEXT: 'TEXT'>, <TokenType.POLYGON: 'POLYGON'>, <TokenType.USERDEFINED: 'USERDEFINED'>, <TokenType.MONEY: 'MONEY'>, <TokenType.ROW: 'ROW'>, <TokenType.TSTZMULTIRANGE: 'TSTZMULTIRANGE'>, <TokenType.VARBINARY: 'VARBINARY'>, <TokenType.OPERATOR: 'OPERATOR'>, <TokenType.PROCEDURE: 'PROCEDURE'>, <TokenType.DATABASE: 'DATABASE'>, <TokenType.TIMESTAMPLTZ: 'TIMESTAMPLTZ'>, <TokenType.ALL: 'ALL'>, <TokenType.UDECIMAL: 'UDECIMAL'>, <TokenType.WAREHOUSE: 'WAREHOUSE'>, <TokenType.EXECUTE: 'EXECUTE'>, <TokenType.XML: 'XML'>, <TokenType.DECIMAL32: 'DECIMAL32'>, <TokenType.FLOAT: 'FLOAT'>, <TokenType.COPY: 'COPY'>, <TokenType.TINYINT: 'TINYINT'>, <TokenType.KILL: 'KILL'>, <TokenType.TINYBLOB: 'TINYBLOB'>, <TokenType.MERGE: 'MERGE'>, <TokenType.OVERLAPS: 'OVERLAPS'>, <TokenType.DECIMAL128: 'DECIMAL128'>, <TokenType.BIGSERIAL: 'BIGSERIAL'>, <TokenType.UBIGINT: 'UBIGINT'>, <TokenType.ENUM16: 'ENUM16'>, <TokenType.CACHE: 'CACHE'>, <TokenType.NUMRANGE: 'NUMRANGE'>, <TokenType.BIGINT: 'BIGINT'>, <TokenType.SEQUENCE: 'SEQUENCE'>, <TokenType.LOWCARDINALITY: 'LOWCARDINALITY'>, <TokenType.IS: 'IS'>, <TokenType.NCHAR: 'NCHAR'>, <TokenType.ROLLUP: 'ROLLUP'>, <TokenType.DECIMAL256: 'DECIMAL256'>, <TokenType.FUNCTION: 'FUNCTION'>, <TokenType.LINESTRING: 'LINESTRING'>, <TokenType.IPV4: 'IPV4'>, <TokenType.STORAGE_INTEGRATION: 'STORAGE_INTEGRATION'>, <TokenType.INET: 'INET'>, <TokenType.RING: 'RING'>, <TokenType.TIME: 'TIME'>, <TokenType.RANGE: 'RANGE'>, <TokenType.FIXEDSTRING: 'FIXEDSTRING'>, <TokenType.FILTER: 'FILTER'>, <TokenType.VAR: 'VAR'>, <TokenType.IPV6: 'IPV6'>, <TokenType.TAG: 'TAG'>, <TokenType.UINT256: 'UINT256'>, <TokenType.CASE: 'CASE'>, <TokenType.TDIGEST: 'TDIGEST'>, <TokenType.SETTINGS: 'SETTINGS'>, <TokenType.YEAR: 'YEAR'>, <TokenType.ANY: 'ANY'>, <TokenType.CURRENT_DATETIME: 'CURRENT_DATETIME'>, <TokenType.USE: 'USE'>, <TokenType.LONGTEXT: 'LONGTEXT'>, <TokenType.DATETIME64: 'DATETIME64'>, <TokenType.COMMENT: 'COMMENT'>, <TokenType.DELETE: 'DELETE'>, <TokenType.ESCAPE: 'ESCAPE'>, <TokenType.END: 'END'>, <TokenType.CHAR: 'CHAR'>, <TokenType.IDENTIFIER: 'IDENTIFIER'>, <TokenType.MULTIPOLYGON: 'MULTIPOLYGON'>, <TokenType.COMMAND: 'COMMAND'>, <TokenType.CUBE: 'CUBE'>, <TokenType.SIMPLEAGGREGATEFUNCTION: 'SIMPLEAGGREGATEFUNCTION'>, <TokenType.INT8MULTIRANGE: 'INT8MULTIRANGE'>, <TokenType.PRAGMA: 'PRAGMA'>, <TokenType.ASC: 'ASC'>, <TokenType.NULL: 'NULL'>, <TokenType.TIMETZ: 'TIMETZ'>, <TokenType.VECTOR: 'VECTOR'>, <TokenType.MEDIUMBLOB: 'MEDIUMBLOB'>, <TokenType.MEDIUMTEXT: 'MEDIUMTEXT'>, <TokenType.UNNEST: 'UNNEST'>}
SHOW_TRIE: Dict = {}
SET_TRIE: Dict = {'GLOBAL': {0: True}, 'LOCAL': {0: True}, 'SESSION': {0: True}, 'TRANSACTION': {0: True}}
Inherited Members
sqlglot.parser.Parser
Parser
NO_PAREN_FUNCTIONS
STRUCT_TYPE_TOKENS
NESTED_TYPE_TOKENS
ENUM_TYPE_TOKENS
AGGREGATE_TYPE_TOKENS
TYPE_TOKENS
SIGNED_TO_UNSIGNED_TYPE_TOKEN
SUBQUERY_PREDICATES
RESERVED_TOKENS
DB_CREATABLES
CREATABLES
ALTERABLES
INTERVAL_VARS
ALIAS_TOKENS
ARRAY_CONSTRUCTORS
COMMENT_TABLE_ALIAS_TOKENS
UPDATE_ALIAS_TOKENS
TRIM_TYPES
FUNC_TOKENS
CONJUNCTION
ASSIGNMENT
DISJUNCTION
EQUALITY
COMPARISON
TERM
FACTOR
TIMES
TIMESTAMPS
SET_OPERATIONS
JOIN_METHODS
JOIN_SIDES
JOIN_KINDS
JOIN_HINTS
LAMBDAS
COLUMN_OPERATORS
EXPRESSION_PARSERS
STATEMENT_PARSERS
UNARY_PARSERS
STRING_PARSERS
NUMERIC_PARSERS
PRIMARY_PARSERS
PROPERTY_PARSERS
CONSTRAINT_PARSERS
ALTER_PARSERS
ALTER_ALTER_PARSERS
SCHEMA_UNNAMED_CONSTRAINTS
INVALID_FUNC_NAME_TOKENS
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
SCHEMA_BINDING_OPTIONS
PROCEDURE_OPTIONS
EXECUTE_AS_OPTIONS
KEY_CONSTRAINT_OPTIONS
INSERT_ALTERNATIVES
CLONE_KEYWORDS
HISTORICAL_DATA_PREFIX
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
NULL_TOKENS
UNNEST_OFFSET_ALIAS_TOKENS
SELECT_START_TOKENS
COPY_INTO_VARLEN_OPTIONS
IS_JSON_PREDICATE_KIND
ODBC_DATETIME_LITERALS
ON_CONDITION_TOKENS
PRIVILEGE_FOLLOW_TOKENS
DESCRIBE_STYLES
OPERATION_MODIFIERS
STRICT_CAST
PREFIXED_PIVOT_COLUMNS
IDENTIFY_PIVOT_STRINGS
LOG_DEFAULTS_TO_LN
ALTER_TABLE_ADD_REQUIRED_FOR_EACH_COLUMN
TABLESAMPLE_CSV
DEFAULT_SAMPLING_METHOD
SET_REQUIRES_ASSIGNMENT_DELIMITER
TRIM_PATTERN_FIRST
STRING_ALIASES
MODIFIERS_ATTACHED_TO_SET_OP
SET_OP_MODIFIERS
NO_PAREN_IF_COMMANDS
JSON_ARROWS_REQUIRE_JSON_TYPE
COLON_IS_VARIANT_EXTRACT
VALUES_FOLLOWED_BY_PAREN
SUPPORTS_IMPLICIT_UNNEST
INTERVAL_SPANS
SUPPORTS_PARTITION_SELECTION
error_level
error_message_context
max_errors
dialect
reset
parse
parse_into
check_errors
raise_error
expression
validate_expression
errors
sql
class DuckDB.Generator(sqlglot.generator.Generator):
486    class Generator(generator.Generator):
487        PARAMETER_TOKEN = "$"
488        NAMED_PLACEHOLDER_TOKEN = "$"
489        JOIN_HINTS = False
490        TABLE_HINTS = False
491        QUERY_HINTS = False
492        LIMIT_FETCH = "LIMIT"
493        STRUCT_DELIMITER = ("(", ")")
494        RENAME_TABLE_WITH_DB = False
495        NVL2_SUPPORTED = False
496        SEMI_ANTI_JOIN_WITH_SIDE = False
497        TABLESAMPLE_KEYWORDS = "USING SAMPLE"
498        TABLESAMPLE_SEED_KEYWORD = "REPEATABLE"
499        LAST_DAY_SUPPORTS_DATE_PART = False
500        JSON_KEY_VALUE_PAIR_SEP = ","
501        IGNORE_NULLS_IN_FUNC = True
502        JSON_PATH_BRACKETED_KEY_SUPPORTED = False
503        SUPPORTS_CREATE_TABLE_LIKE = False
504        MULTI_ARG_DISTINCT = False
505        CAN_IMPLEMENT_ARRAY_ANY = True
506        SUPPORTS_TO_NUMBER = False
507        COPY_HAS_INTO_KEYWORD = False
508        STAR_EXCEPT = "EXCLUDE"
509        PAD_FILL_PATTERN_IS_REQUIRED = True
510        ARRAY_CONCAT_IS_VAR_LEN = False
511        ARRAY_SIZE_DIM_REQUIRED = False
512
513        TRANSFORMS = {
514            **generator.Generator.TRANSFORMS,
515            exp.ApproxDistinct: approx_count_distinct_sql,
516            exp.Array: inline_array_unless_query,
517            exp.ArrayFilter: rename_func("LIST_FILTER"),
518            exp.ArgMax: arg_max_or_min_no_count("ARG_MAX"),
519            exp.ArgMin: arg_max_or_min_no_count("ARG_MIN"),
520            exp.ArraySort: _array_sort_sql,
521            exp.ArraySum: rename_func("LIST_SUM"),
522            exp.BitwiseXor: rename_func("XOR"),
523            exp.CommentColumnConstraint: no_comment_column_constraint_sql,
524            exp.CurrentDate: lambda *_: "CURRENT_DATE",
525            exp.CurrentTime: lambda *_: "CURRENT_TIME",
526            exp.CurrentTimestamp: lambda *_: "CURRENT_TIMESTAMP",
527            exp.DayOfMonth: rename_func("DAYOFMONTH"),
528            exp.DayOfWeek: rename_func("DAYOFWEEK"),
529            exp.DayOfWeekIso: rename_func("ISODOW"),
530            exp.DayOfYear: rename_func("DAYOFYEAR"),
531            exp.DataType: _datatype_sql,
532            exp.Date: _date_sql,
533            exp.DateAdd: _date_delta_sql,
534            exp.DateFromParts: rename_func("MAKE_DATE"),
535            exp.DateSub: _date_delta_sql,
536            exp.DateDiff: _date_diff_sql,
537            exp.DateStrToDate: datestrtodate_sql,
538            exp.Datetime: no_datetime_sql,
539            exp.DatetimeSub: _date_delta_sql,
540            exp.DatetimeAdd: _date_delta_sql,
541            exp.DateToDi: lambda self,
542            e: f"CAST(STRFTIME({self.sql(e, 'this')}, {DuckDB.DATEINT_FORMAT}) AS INT)",
543            exp.Decode: lambda self, e: encode_decode_sql(self, e, "DECODE", replace=False),
544            exp.DiToDate: lambda self,
545            e: f"CAST(STRPTIME(CAST({self.sql(e, 'this')} AS TEXT), {DuckDB.DATEINT_FORMAT}) AS DATE)",
546            exp.Encode: lambda self, e: encode_decode_sql(self, e, "ENCODE", replace=False),
547            exp.GenerateDateArray: _generate_datetime_array_sql,
548            exp.GenerateTimestampArray: _generate_datetime_array_sql,
549            exp.Explode: rename_func("UNNEST"),
550            exp.IntDiv: lambda self, e: self.binary(e, "//"),
551            exp.IsInf: rename_func("ISINF"),
552            exp.IsNan: rename_func("ISNAN"),
553            exp.JSONBExists: rename_func("JSON_EXISTS"),
554            exp.JSONExtract: _arrow_json_extract_sql,
555            exp.JSONExtractArray: _json_extract_value_array_sql,
556            exp.JSONExtractScalar: _arrow_json_extract_sql,
557            exp.JSONFormat: _json_format_sql,
558            exp.JSONValueArray: _json_extract_value_array_sql,
559            exp.Lateral: explode_to_unnest_sql,
560            exp.LogicalOr: rename_func("BOOL_OR"),
561            exp.LogicalAnd: rename_func("BOOL_AND"),
562            exp.MD5Digest: lambda self, e: self.func("UNHEX", self.func("MD5", e.this)),
563            exp.MonthsBetween: lambda self, e: self.func(
564                "DATEDIFF",
565                "'month'",
566                exp.cast(e.expression, exp.DataType.Type.TIMESTAMP, copy=True),
567                exp.cast(e.this, exp.DataType.Type.TIMESTAMP, copy=True),
568            ),
569            exp.PercentileCont: rename_func("QUANTILE_CONT"),
570            exp.PercentileDisc: rename_func("QUANTILE_DISC"),
571            # DuckDB doesn't allow qualified columns inside of PIVOT expressions.
572            # See: https://github.com/duckdb/duckdb/blob/671faf92411182f81dce42ac43de8bfb05d9909e/src/planner/binder/tableref/bind_pivot.cpp#L61-L62
573            exp.Pivot: transforms.preprocess([transforms.unqualify_columns]),
574            exp.RegexpReplace: lambda self, e: self.func(
575                "REGEXP_REPLACE",
576                e.this,
577                e.expression,
578                e.args.get("replacement"),
579                e.args.get("modifiers"),
580            ),
581            exp.RegexpLike: rename_func("REGEXP_MATCHES"),
582            exp.RegexpSplit: rename_func("STR_SPLIT_REGEX"),
583            exp.Return: lambda self, e: self.sql(e, "this"),
584            exp.ReturnsProperty: lambda self, e: "TABLE" if isinstance(e.this, exp.Schema) else "",
585            exp.Rand: rename_func("RANDOM"),
586            exp.SafeDivide: no_safe_divide_sql,
587            exp.SHA: rename_func("SHA1"),
588            exp.SHA2: sha256_sql,
589            exp.Split: rename_func("STR_SPLIT"),
590            exp.SortArray: _sort_array_sql,
591            exp.StrPosition: str_position_sql,
592            exp.StrToUnix: lambda self, e: self.func(
593                "EPOCH", self.func("STRPTIME", e.this, self.format_time(e))
594            ),
595            exp.Struct: _struct_sql,
596            exp.Transform: rename_func("LIST_TRANSFORM"),
597            exp.TimeAdd: _date_delta_sql,
598            exp.Time: no_time_sql,
599            exp.TimeDiff: _timediff_sql,
600            exp.Timestamp: no_timestamp_sql,
601            exp.TimestampDiff: lambda self, e: self.func(
602                "DATE_DIFF", exp.Literal.string(e.unit), e.expression, e.this
603            ),
604            exp.TimestampTrunc: timestamptrunc_sql(),
605            exp.TimeStrToDate: lambda self, e: self.sql(exp.cast(e.this, exp.DataType.Type.DATE)),
606            exp.TimeStrToTime: timestrtotime_sql,
607            exp.TimeStrToUnix: lambda self, e: self.func(
608                "EPOCH", exp.cast(e.this, exp.DataType.Type.TIMESTAMP)
609            ),
610            exp.TimeToStr: lambda self, e: self.func("STRFTIME", e.this, self.format_time(e)),
611            exp.TimeToUnix: rename_func("EPOCH"),
612            exp.TsOrDiToDi: lambda self,
613            e: f"CAST(SUBSTR(REPLACE(CAST({self.sql(e, 'this')} AS TEXT), '-', ''), 1, 8) AS INT)",
614            exp.TsOrDsAdd: _date_delta_sql,
615            exp.TsOrDsDiff: lambda self, e: self.func(
616                "DATE_DIFF",
617                f"'{e.args.get('unit') or 'DAY'}'",
618                exp.cast(e.expression, exp.DataType.Type.TIMESTAMP),
619                exp.cast(e.this, exp.DataType.Type.TIMESTAMP),
620            ),
621            exp.UnixToStr: lambda self, e: self.func(
622                "STRFTIME", self.func("TO_TIMESTAMP", e.this), self.format_time(e)
623            ),
624            exp.DatetimeTrunc: lambda self, e: self.func(
625                "DATE_TRUNC", unit_to_str(e), exp.cast(e.this, exp.DataType.Type.DATETIME)
626            ),
627            exp.UnixToTime: _unix_to_time_sql,
628            exp.UnixToTimeStr: lambda self, e: f"CAST(TO_TIMESTAMP({self.sql(e, 'this')}) AS TEXT)",
629            exp.VariancePop: rename_func("VAR_POP"),
630            exp.WeekOfYear: rename_func("WEEKOFYEAR"),
631            exp.Xor: bool_xor_sql,
632            exp.Levenshtein: unsupported_args("ins_cost", "del_cost", "sub_cost", "max_dist")(
633                rename_func("LEVENSHTEIN")
634            ),
635        }
636
637        SUPPORTED_JSON_PATH_PARTS = {
638            exp.JSONPathKey,
639            exp.JSONPathRoot,
640            exp.JSONPathSubscript,
641            exp.JSONPathWildcard,
642        }
643
644        TYPE_MAPPING = {
645            **generator.Generator.TYPE_MAPPING,
646            exp.DataType.Type.BINARY: "BLOB",
647            exp.DataType.Type.BPCHAR: "TEXT",
648            exp.DataType.Type.CHAR: "TEXT",
649            exp.DataType.Type.FLOAT: "REAL",
650            exp.DataType.Type.NCHAR: "TEXT",
651            exp.DataType.Type.NVARCHAR: "TEXT",
652            exp.DataType.Type.UINT: "UINTEGER",
653            exp.DataType.Type.VARBINARY: "BLOB",
654            exp.DataType.Type.ROWVERSION: "BLOB",
655            exp.DataType.Type.VARCHAR: "TEXT",
656            exp.DataType.Type.TIMESTAMPNTZ: "TIMESTAMP",
657            exp.DataType.Type.TIMESTAMP_S: "TIMESTAMP_S",
658            exp.DataType.Type.TIMESTAMP_MS: "TIMESTAMP_MS",
659            exp.DataType.Type.TIMESTAMP_NS: "TIMESTAMP_NS",
660        }
661
662        # https://github.com/duckdb/duckdb/blob/ff7f24fd8e3128d94371827523dae85ebaf58713/third_party/libpg_query/grammar/keywords/reserved_keywords.list#L1-L77
663        RESERVED_KEYWORDS = {
664            "array",
665            "analyse",
666            "union",
667            "all",
668            "when",
669            "in_p",
670            "default",
671            "create_p",
672            "window",
673            "asymmetric",
674            "to",
675            "else",
676            "localtime",
677            "from",
678            "end_p",
679            "select",
680            "current_date",
681            "foreign",
682            "with",
683            "grant",
684            "session_user",
685            "or",
686            "except",
687            "references",
688            "fetch",
689            "limit",
690            "group_p",
691            "leading",
692            "into",
693            "collate",
694            "offset",
695            "do",
696            "then",
697            "localtimestamp",
698            "check_p",
699            "lateral_p",
700            "current_role",
701            "where",
702            "asc_p",
703            "placing",
704            "desc_p",
705            "user",
706            "unique",
707            "initially",
708            "column",
709            "both",
710            "some",
711            "as",
712            "any",
713            "only",
714            "deferrable",
715            "null_p",
716            "current_time",
717            "true_p",
718            "table",
719            "case",
720            "trailing",
721            "variadic",
722            "for",
723            "on",
724            "distinct",
725            "false_p",
726            "not",
727            "constraint",
728            "current_timestamp",
729            "returning",
730            "primary",
731            "intersect",
732            "having",
733            "analyze",
734            "current_user",
735            "and",
736            "cast",
737            "symmetric",
738            "using",
739            "order",
740            "current_catalog",
741        }
742
743        UNWRAPPED_INTERVAL_VALUES = (exp.Literal, exp.Paren)
744
745        # DuckDB doesn't generally support CREATE TABLE .. properties
746        # https://duckdb.org/docs/sql/statements/create_table.html
747        PROPERTIES_LOCATION = {
748            prop: exp.Properties.Location.UNSUPPORTED
749            for prop in generator.Generator.PROPERTIES_LOCATION
750        }
751
752        # There are a few exceptions (e.g. temporary tables) which are supported or
753        # can be transpiled to DuckDB, so we explicitly override them accordingly
754        PROPERTIES_LOCATION[exp.LikeProperty] = exp.Properties.Location.POST_SCHEMA
755        PROPERTIES_LOCATION[exp.TemporaryProperty] = exp.Properties.Location.POST_CREATE
756        PROPERTIES_LOCATION[exp.ReturnsProperty] = exp.Properties.Location.POST_ALIAS
757
758        def fromiso8601timestamp_sql(self, expression: exp.FromISO8601Timestamp) -> str:
759            return self.sql(exp.cast(expression.this, exp.DataType.Type.TIMESTAMPTZ))
760
761        def strtotime_sql(self, expression: exp.StrToTime) -> str:
762            if expression.args.get("safe"):
763                formatted_time = self.format_time(expression)
764                return f"CAST({self.func('TRY_STRPTIME', expression.this, formatted_time)} AS TIMESTAMP)"
765            return str_to_time_sql(self, expression)
766
767        def strtodate_sql(self, expression: exp.StrToDate) -> str:
768            if expression.args.get("safe"):
769                formatted_time = self.format_time(expression)
770                return f"CAST({self.func('TRY_STRPTIME', expression.this, formatted_time)} AS DATE)"
771            return f"CAST({str_to_time_sql(self, expression)} AS DATE)"
772
773        def parsejson_sql(self, expression: exp.ParseJSON) -> str:
774            arg = expression.this
775            if expression.args.get("safe"):
776                return self.sql(exp.case().when(exp.func("json_valid", arg), arg).else_(exp.null()))
777            return self.func("JSON", arg)
778
779        def timefromparts_sql(self, expression: exp.TimeFromParts) -> str:
780            nano = expression.args.get("nano")
781            if nano is not None:
782                expression.set(
783                    "sec", expression.args["sec"] + nano.pop() / exp.Literal.number(1000000000.0)
784                )
785
786            return rename_func("MAKE_TIME")(self, expression)
787
788        def timestampfromparts_sql(self, expression: exp.TimestampFromParts) -> str:
789            sec = expression.args["sec"]
790
791            milli = expression.args.get("milli")
792            if milli is not None:
793                sec += milli.pop() / exp.Literal.number(1000.0)
794
795            nano = expression.args.get("nano")
796            if nano is not None:
797                sec += nano.pop() / exp.Literal.number(1000000000.0)
798
799            if milli or nano:
800                expression.set("sec", sec)
801
802            return rename_func("MAKE_TIMESTAMP")(self, expression)
803
804        def tablesample_sql(
805            self,
806            expression: exp.TableSample,
807            tablesample_keyword: t.Optional[str] = None,
808        ) -> str:
809            if not isinstance(expression.parent, exp.Select):
810                # This sample clause only applies to a single source, not the entire resulting relation
811                tablesample_keyword = "TABLESAMPLE"
812
813            if expression.args.get("size"):
814                method = expression.args.get("method")
815                if method and method.name.upper() != "RESERVOIR":
816                    self.unsupported(
817                        f"Sampling method {method} is not supported with a discrete sample count, "
818                        "defaulting to reservoir sampling"
819                    )
820                    expression.set("method", exp.var("RESERVOIR"))
821
822            return super().tablesample_sql(expression, tablesample_keyword=tablesample_keyword)
823
824        def interval_sql(self, expression: exp.Interval) -> str:
825            multiplier: t.Optional[int] = None
826            unit = expression.text("unit").lower()
827
828            if unit.startswith("week"):
829                multiplier = 7
830            if unit.startswith("quarter"):
831                multiplier = 90
832
833            if multiplier:
834                return f"({multiplier} * {super().interval_sql(exp.Interval(this=expression.this, unit=exp.var('DAY')))})"
835
836            return super().interval_sql(expression)
837
838        def columndef_sql(self, expression: exp.ColumnDef, sep: str = " ") -> str:
839            if isinstance(expression.parent, exp.UserDefinedFunction):
840                return self.sql(expression, "this")
841            return super().columndef_sql(expression, sep)
842
843        def join_sql(self, expression: exp.Join) -> str:
844            if (
845                expression.side == "LEFT"
846                and not expression.args.get("on")
847                and isinstance(expression.this, exp.Unnest)
848            ):
849                # Some dialects support `LEFT JOIN UNNEST(...)` without an explicit ON clause
850                # DuckDB doesn't, but we can just add a dummy ON clause that is always true
851                return super().join_sql(expression.on(exp.true()))
852
853            return super().join_sql(expression)
854
855        def generateseries_sql(self, expression: exp.GenerateSeries) -> str:
856            # GENERATE_SERIES(a, b) -> [a, b], RANGE(a, b) -> [a, b)
857            if expression.args.get("is_end_exclusive"):
858                return rename_func("RANGE")(self, expression)
859
860            return self.function_fallback_sql(expression)
861
862        def bracket_sql(self, expression: exp.Bracket) -> str:
863            this = expression.this
864            if isinstance(this, exp.Array):
865                this.replace(exp.paren(this))
866
867            bracket = super().bracket_sql(expression)
868
869            if not expression.args.get("returns_list_for_maps"):
870                if not this.type:
871                    from sqlglot.optimizer.annotate_types import annotate_types
872
873                    this = annotate_types(this)
874
875                if this.is_type(exp.DataType.Type.MAP):
876                    bracket = f"({bracket})[1]"
877
878            return bracket
879
880        def withingroup_sql(self, expression: exp.WithinGroup) -> str:
881            expression_sql = self.sql(expression, "expression")
882
883            func = expression.this
884            if isinstance(func, exp.PERCENTILES):
885                # Make the order key the first arg and slide the fraction to the right
886                # https://duckdb.org/docs/sql/aggregates#ordered-set-aggregate-functions
887                order_col = expression.find(exp.Ordered)
888                if order_col:
889                    func.set("expression", func.this)
890                    func.set("this", order_col.this)
891
892            this = self.sql(expression, "this").rstrip(")")
893
894            return f"{this}{expression_sql})"
895
896        def length_sql(self, expression: exp.Length) -> str:
897            arg = expression.this
898
899            # Dialects like BQ and Snowflake also accept binary values as args, so
900            # DDB will attempt to infer the type or resort to case/when resolution
901            if not expression.args.get("binary") or arg.is_string:
902                return self.func("LENGTH", arg)
903
904            if not arg.type:
905                from sqlglot.optimizer.annotate_types import annotate_types
906
907                arg = annotate_types(arg)
908
909            if arg.is_type(*exp.DataType.TEXT_TYPES):
910                return self.func("LENGTH", arg)
911
912            # We need these casts to make duckdb's static type checker happy
913            blob = exp.cast(arg, exp.DataType.Type.VARBINARY)
914            varchar = exp.cast(arg, exp.DataType.Type.VARCHAR)
915
916            case = (
917                exp.case(self.func("TYPEOF", arg))
918                .when(
919                    "'VARCHAR'", exp.Anonymous(this="LENGTH", expressions=[varchar])
920                )  # anonymous to break length_sql recursion
921                .when("'BLOB'", self.func("OCTET_LENGTH", blob))
922            )
923
924            return self.sql(case)
925
926        def objectinsert_sql(self, expression: exp.ObjectInsert) -> str:
927            this = expression.this
928            key = expression.args.get("key")
929            key_sql = key.name if isinstance(key, exp.Expression) else ""
930            value_sql = self.sql(expression, "value")
931
932            kv_sql = f"{key_sql} := {value_sql}"
933
934            # If the input struct is empty e.g. transpiling OBJECT_INSERT(OBJECT_CONSTRUCT(), key, value) from Snowflake
935            # then we can generate STRUCT_PACK which will build it since STRUCT_INSERT({}, key := value) is not valid DuckDB
936            if isinstance(this, exp.Struct) and not this.expressions:
937                return self.func("STRUCT_PACK", kv_sql)
938
939            return self.func("STRUCT_INSERT", this, kv_sql)
940
941        def unnest_sql(self, expression: exp.Unnest) -> str:
942            explode_array = expression.args.get("explode_array")
943            if explode_array:
944                # In BigQuery, UNNESTing a nested array leads to explosion of the top-level array & struct
945                # This is transpiled to DDB by transforming "FROM UNNEST(...)" to "FROM (SELECT UNNEST(..., max_depth => 2))"
946                expression.expressions.append(
947                    exp.Kwarg(this=exp.var("max_depth"), expression=exp.Literal.number(2))
948                )
949
950                # If BQ's UNNEST is aliased, we transform it from a column alias to a table alias in DDB
951                alias = expression.args.get("alias")
952                if alias:
953                    expression.set("alias", None)
954                    alias = exp.TableAlias(this=seq_get(alias.args.get("columns"), 0))
955
956                unnest_sql = super().unnest_sql(expression)
957                select = exp.Select(expressions=[unnest_sql]).subquery(alias)
958                return self.sql(select)
959
960            return super().unnest_sql(expression)
961
962        def ignorenulls_sql(self, expression: exp.IgnoreNulls) -> str:
963            if isinstance(expression.this, WINDOW_FUNCS_WITH_IGNORE_NULLS):
964                # DuckDB should render IGNORE NULLS only for the general-purpose
965                # window functions that accept it e.g. FIRST_VALUE(... IGNORE NULLS) OVER (...)
966                return super().ignorenulls_sql(expression)
967
968            return self.sql(expression, "this")
969
970        def arraytostring_sql(self, expression: exp.ArrayToString) -> str:
971            this = self.sql(expression, "this")
972            null_text = self.sql(expression, "null")
973
974            if null_text:
975                this = f"LIST_TRANSFORM({this}, x -> COALESCE(x, {null_text}))"
976
977            return self.func("ARRAY_TO_STRING", this, expression.expression)
978
979        @unsupported_args("position", "occurrence")
980        def regexpextract_sql(self, expression: exp.RegexpExtract) -> str:
981            group = expression.args.get("group")
982            params = expression.args.get("parameters")
983
984            # Do not render group if there is no following argument,
985            # and it's the default value for this dialect
986            if (
987                not params
988                and group
989                and group.name == str(self.dialect.REGEXP_EXTRACT_DEFAULT_GROUP)
990            ):
991                group = None
992            return self.func(
993                "REGEXP_EXTRACT", expression.this, expression.expression, group, params
994            )

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
PARAMETER_TOKEN = '$'
NAMED_PLACEHOLDER_TOKEN = '$'
JOIN_HINTS = False
TABLE_HINTS = False
QUERY_HINTS = False
LIMIT_FETCH = 'LIMIT'
STRUCT_DELIMITER = ('(', ')')
RENAME_TABLE_WITH_DB = False
NVL2_SUPPORTED = False
SEMI_ANTI_JOIN_WITH_SIDE = False
TABLESAMPLE_KEYWORDS = 'USING SAMPLE'
TABLESAMPLE_SEED_KEYWORD = 'REPEATABLE'
LAST_DAY_SUPPORTS_DATE_PART = False
JSON_KEY_VALUE_PAIR_SEP = ','
IGNORE_NULLS_IN_FUNC = True
JSON_PATH_BRACKETED_KEY_SUPPORTED = False
SUPPORTS_CREATE_TABLE_LIKE = False
MULTI_ARG_DISTINCT = False
CAN_IMPLEMENT_ARRAY_ANY = True
SUPPORTS_TO_NUMBER = False
COPY_HAS_INTO_KEYWORD = False
STAR_EXCEPT = 'EXCLUDE'
PAD_FILL_PATTERN_IS_REQUIRED = True
ARRAY_CONCAT_IS_VAR_LEN = False
ARRAY_SIZE_DIM_REQUIRED = False
TRANSFORMS = {<class 'sqlglot.expressions.JSONPathKey'>: <function <lambda>>, <class 'sqlglot.expressions.JSONPathRoot'>: <function <lambda>>, <class 'sqlglot.expressions.JSONPathSubscript'>: <function <lambda>>, <class 'sqlglot.expressions.JSONPathWildcard'>: <function <lambda>>, <class 'sqlglot.expressions.AllowedValuesProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ArrayContainsAll'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ArrayOverlaps'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.AutoRefreshProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.BackupProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.CaseSpecificColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.CharacterSetColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.CharacterSetProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ClusteredColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.CollateColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.CommentColumnConstraint'>: <function no_comment_column_constraint_sql>, <class 'sqlglot.expressions.ConnectByRoot'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.CopyGrantsProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.DateFormatColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.DefaultColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.DynamicProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.EmptyProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.EncodeColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.EphemeralColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ExcludeColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ExecuteAsProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.Except'>: <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.Intersect'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.IntervalSpan'>: <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.Operator'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.OutputModelProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.PathColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.PivotAny'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ProjectionPolicyColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.RemoteWithConnectionModelProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ReturnsProperty'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.SampleProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SecureProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SecurityProperty'>: <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 Generator.<lambda>>, <class 'sqlglot.expressions.Stream'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.StreamingTableProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.StrictProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SwapTable'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.TemporaryProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.TagColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.TitleColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.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.Union'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.UnloggedProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.Uuid'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.UppercaseColumnConstraint'>: <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.WithProcedureOptions'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.WithSchemaBindingProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.WithOperator'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ApproxDistinct'>: <function approx_count_distinct_sql>, <class 'sqlglot.expressions.Array'>: <function inline_array_unless_query>, <class 'sqlglot.expressions.ArrayFilter'>: <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.ArraySort'>: <function _array_sort_sql>, <class 'sqlglot.expressions.ArraySum'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.BitwiseXor'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.CurrentDate'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.CurrentTime'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.CurrentTimestamp'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.DayOfMonth'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.DayOfWeek'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.DayOfWeekIso'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.DayOfYear'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.DataType'>: <function _datatype_sql>, <class 'sqlglot.expressions.Date'>: <function _date_sql>, <class 'sqlglot.expressions.DateAdd'>: <function _date_delta_sql>, <class 'sqlglot.expressions.DateFromParts'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.DateSub'>: <function _date_delta_sql>, <class 'sqlglot.expressions.DateDiff'>: <function _date_diff_sql>, <class 'sqlglot.expressions.DateStrToDate'>: <function datestrtodate_sql>, <class 'sqlglot.expressions.Datetime'>: <function no_datetime_sql>, <class 'sqlglot.expressions.DatetimeSub'>: <function _date_delta_sql>, <class 'sqlglot.expressions.DatetimeAdd'>: <function _date_delta_sql>, <class 'sqlglot.expressions.DateToDi'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.Decode'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.DiToDate'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.Encode'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.GenerateDateArray'>: <function _generate_datetime_array_sql>, <class 'sqlglot.expressions.GenerateTimestampArray'>: <function _generate_datetime_array_sql>, <class 'sqlglot.expressions.Explode'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.IntDiv'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.IsInf'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.IsNan'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.JSONBExists'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.JSONExtract'>: <function _arrow_json_extract_sql>, <class 'sqlglot.expressions.JSONExtractArray'>: <function _json_extract_value_array_sql>, <class 'sqlglot.expressions.JSONExtractScalar'>: <function _arrow_json_extract_sql>, <class 'sqlglot.expressions.JSONFormat'>: <function _json_format_sql>, <class 'sqlglot.expressions.JSONValueArray'>: <function _json_extract_value_array_sql>, <class 'sqlglot.expressions.Lateral'>: <function explode_to_unnest_sql>, <class 'sqlglot.expressions.LogicalOr'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.LogicalAnd'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.MD5Digest'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.MonthsBetween'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.PercentileCont'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.PercentileDisc'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.Pivot'>: <function preprocess.<locals>._to_sql>, <class 'sqlglot.expressions.RegexpReplace'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.RegexpLike'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.RegexpSplit'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.Return'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.Rand'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.SafeDivide'>: <function no_safe_divide_sql>, <class 'sqlglot.expressions.SHA'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.SHA2'>: <function sha256_sql>, <class 'sqlglot.expressions.Split'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.SortArray'>: <function _sort_array_sql>, <class 'sqlglot.expressions.StrPosition'>: <function str_position_sql>, <class 'sqlglot.expressions.StrToUnix'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.Struct'>: <function _struct_sql>, <class 'sqlglot.expressions.Transform'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.TimeAdd'>: <function _date_delta_sql>, <class 'sqlglot.expressions.Time'>: <function no_time_sql>, <class 'sqlglot.expressions.TimeDiff'>: <function _timediff_sql>, <class 'sqlglot.expressions.Timestamp'>: <function no_timestamp_sql>, <class 'sqlglot.expressions.TimestampDiff'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.TimestampTrunc'>: <function timestamptrunc_sql.<locals>._timestamptrunc_sql>, <class 'sqlglot.expressions.TimeStrToDate'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.TimeStrToTime'>: <function timestrtotime_sql>, <class 'sqlglot.expressions.TimeStrToUnix'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.TimeToStr'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.TimeToUnix'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.TsOrDiToDi'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.TsOrDsAdd'>: <function _date_delta_sql>, <class 'sqlglot.expressions.TsOrDsDiff'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.UnixToStr'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.DatetimeTrunc'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.UnixToTime'>: <function _unix_to_time_sql>, <class 'sqlglot.expressions.UnixToTimeStr'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.VariancePop'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.WeekOfYear'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.Xor'>: <function bool_xor_sql>, <class 'sqlglot.expressions.Levenshtein'>: <function rename_func.<locals>.<lambda>>}
TYPE_MAPPING = {<Type.NCHAR: 'NCHAR'>: 'TEXT', <Type.NVARCHAR: 'NVARCHAR'>: 'TEXT', <Type.MEDIUMTEXT: 'MEDIUMTEXT'>: 'TEXT', <Type.LONGTEXT: 'LONGTEXT'>: 'TEXT', <Type.TINYTEXT: 'TINYTEXT'>: 'TEXT', <Type.MEDIUMBLOB: 'MEDIUMBLOB'>: 'BLOB', <Type.LONGBLOB: 'LONGBLOB'>: 'BLOB', <Type.TINYBLOB: 'TINYBLOB'>: 'BLOB', <Type.INET: 'INET'>: 'INET', <Type.ROWVERSION: 'ROWVERSION'>: 'BLOB', <Type.BINARY: 'BINARY'>: 'BLOB', <Type.BPCHAR: 'BPCHAR'>: 'TEXT', <Type.CHAR: 'CHAR'>: 'TEXT', <Type.FLOAT: 'FLOAT'>: 'REAL', <Type.UINT: 'UINT'>: 'UINTEGER', <Type.VARBINARY: 'VARBINARY'>: 'BLOB', <Type.VARCHAR: 'VARCHAR'>: 'TEXT', <Type.TIMESTAMPNTZ: 'TIMESTAMPNTZ'>: 'TIMESTAMP', <Type.TIMESTAMP_S: 'TIMESTAMP_S'>: 'TIMESTAMP_S', <Type.TIMESTAMP_MS: 'TIMESTAMP_MS'>: 'TIMESTAMP_MS', <Type.TIMESTAMP_NS: 'TIMESTAMP_NS'>: 'TIMESTAMP_NS'}
RESERVED_KEYWORDS = {'from', 'fetch', 'into', 'with', 'else', 'current_timestamp', 'current_date', 'constraint', 'table', 'some', 'null_p', 'default', 'primary', 'except', 'using', 'user', 'for', 'analyse', 'all', 'false_p', 'then', 'unique', 'select', 'create_p', 'limit', 'having', 'localtime', 'session_user', 'lateral_p', 'deferrable', 'distinct', 'in_p', 'asc_p', 'check_p', 'analyze', 'offset', 'initially', 'column', 'both', 'leading', 'current_role', 'returning', 'desc_p', 'do', 'asymmetric', 'current_user', 'symmetric', 'foreign', 'to', 'or', 'variadic', 'not', 'where', 'and', 'true_p', 'case', 'trailing', 'group_p', 'current_time', 'order', 'collate', 'cast', 'any', 'references', 'localtimestamp', 'placing', 'only', 'array', 'as', 'current_catalog', 'end_p', 'when', 'union', 'grant', 'window', 'on', 'intersect'}
UNWRAPPED_INTERVAL_VALUES = (<class 'sqlglot.expressions.Literal'>, <class 'sqlglot.expressions.Paren'>)
PROPERTIES_LOCATION = {<class 'sqlglot.expressions.AllowedValuesProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.AlgorithmProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.AutoIncrementProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.AutoRefreshProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.BackupProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.BlockCompressionProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.CharacterSetProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.ChecksumProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.CollateProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.CopyGrantsProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.Cluster'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.ClusteredByProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.DistributedByProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.DuplicateKeyProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.DataBlocksizeProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.DataDeletionProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.DefinerProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.DictRange'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.DictProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.DynamicProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.DistKeyProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.DistStyleProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.EmptyProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.EngineProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.ExecuteAsProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.ExternalProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.FallbackProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.FileFormatProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.FreespaceProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.GlobalProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.HeapProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.InheritsProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.IcebergProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.InputModelProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.IsolatedLoadingProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.JournalProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.LanguageProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.LikeProperty'>: <Location.POST_SCHEMA: 'POST_SCHEMA'>, <class 'sqlglot.expressions.LocationProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.LockProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.LockingProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.LogProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.MaterializedProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.MergeBlockRatioProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.NoPrimaryIndexProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.OnProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.OnCommitProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.Order'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.OutputModelProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.PartitionedByProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.PartitionedOfProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.PrimaryKey'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.Property'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.RemoteWithConnectionModelProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.ReturnsProperty'>: <Location.POST_ALIAS: 'POST_ALIAS'>, <class 'sqlglot.expressions.RowFormatProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.RowFormatDelimitedProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.RowFormatSerdeProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.SampleProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.SchemaCommentProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.SecureProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.SecurityProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.SerdeProperties'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.Set'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.SettingsProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.SetProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.SetConfigProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.SharingProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.SequenceProperties'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.SortKeyProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.SqlReadWriteProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.SqlSecurityProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.StabilityProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.StreamingTableProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.StrictProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.TemporaryProperty'>: <Location.POST_CREATE: 'POST_CREATE'>, <class 'sqlglot.expressions.ToTableProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.TransientProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.TransformModelProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.MergeTreeTTL'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.UnloggedProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.ViewAttributeProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.VolatileProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.WithDataProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.WithJournalTableProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.WithProcedureOptions'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.WithSchemaBindingProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.WithSystemVersioningProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>}
def fromiso8601timestamp_sql(self, expression: sqlglot.expressions.FromISO8601Timestamp) -> str:
758        def fromiso8601timestamp_sql(self, expression: exp.FromISO8601Timestamp) -> str:
759            return self.sql(exp.cast(expression.this, exp.DataType.Type.TIMESTAMPTZ))
def strtotime_sql(self, expression: sqlglot.expressions.StrToTime) -> str:
761        def strtotime_sql(self, expression: exp.StrToTime) -> str:
762            if expression.args.get("safe"):
763                formatted_time = self.format_time(expression)
764                return f"CAST({self.func('TRY_STRPTIME', expression.this, formatted_time)} AS TIMESTAMP)"
765            return str_to_time_sql(self, expression)
def strtodate_sql(self, expression: sqlglot.expressions.StrToDate) -> str:
767        def strtodate_sql(self, expression: exp.StrToDate) -> str:
768            if expression.args.get("safe"):
769                formatted_time = self.format_time(expression)
770                return f"CAST({self.func('TRY_STRPTIME', expression.this, formatted_time)} AS DATE)"
771            return f"CAST({str_to_time_sql(self, expression)} AS DATE)"
def parsejson_sql(self, expression: sqlglot.expressions.ParseJSON) -> str:
773        def parsejson_sql(self, expression: exp.ParseJSON) -> str:
774            arg = expression.this
775            if expression.args.get("safe"):
776                return self.sql(exp.case().when(exp.func("json_valid", arg), arg).else_(exp.null()))
777            return self.func("JSON", arg)
def timefromparts_sql(self, expression: sqlglot.expressions.TimeFromParts) -> str:
779        def timefromparts_sql(self, expression: exp.TimeFromParts) -> str:
780            nano = expression.args.get("nano")
781            if nano is not None:
782                expression.set(
783                    "sec", expression.args["sec"] + nano.pop() / exp.Literal.number(1000000000.0)
784                )
785
786            return rename_func("MAKE_TIME")(self, expression)
def timestampfromparts_sql(self, expression: sqlglot.expressions.TimestampFromParts) -> str:
788        def timestampfromparts_sql(self, expression: exp.TimestampFromParts) -> str:
789            sec = expression.args["sec"]
790
791            milli = expression.args.get("milli")
792            if milli is not None:
793                sec += milli.pop() / exp.Literal.number(1000.0)
794
795            nano = expression.args.get("nano")
796            if nano is not None:
797                sec += nano.pop() / exp.Literal.number(1000000000.0)
798
799            if milli or nano:
800                expression.set("sec", sec)
801
802            return rename_func("MAKE_TIMESTAMP")(self, expression)
def tablesample_sql( self, expression: sqlglot.expressions.TableSample, tablesample_keyword: Optional[str] = None) -> str:
804        def tablesample_sql(
805            self,
806            expression: exp.TableSample,
807            tablesample_keyword: t.Optional[str] = None,
808        ) -> str:
809            if not isinstance(expression.parent, exp.Select):
810                # This sample clause only applies to a single source, not the entire resulting relation
811                tablesample_keyword = "TABLESAMPLE"
812
813            if expression.args.get("size"):
814                method = expression.args.get("method")
815                if method and method.name.upper() != "RESERVOIR":
816                    self.unsupported(
817                        f"Sampling method {method} is not supported with a discrete sample count, "
818                        "defaulting to reservoir sampling"
819                    )
820                    expression.set("method", exp.var("RESERVOIR"))
821
822            return super().tablesample_sql(expression, tablesample_keyword=tablesample_keyword)
def interval_sql(self, expression: sqlglot.expressions.Interval) -> str:
824        def interval_sql(self, expression: exp.Interval) -> str:
825            multiplier: t.Optional[int] = None
826            unit = expression.text("unit").lower()
827
828            if unit.startswith("week"):
829                multiplier = 7
830            if unit.startswith("quarter"):
831                multiplier = 90
832
833            if multiplier:
834                return f"({multiplier} * {super().interval_sql(exp.Interval(this=expression.this, unit=exp.var('DAY')))})"
835
836            return super().interval_sql(expression)
def columndef_sql(self, expression: sqlglot.expressions.ColumnDef, sep: str = ' ') -> str:
838        def columndef_sql(self, expression: exp.ColumnDef, sep: str = " ") -> str:
839            if isinstance(expression.parent, exp.UserDefinedFunction):
840                return self.sql(expression, "this")
841            return super().columndef_sql(expression, sep)
def join_sql(self, expression: sqlglot.expressions.Join) -> str:
843        def join_sql(self, expression: exp.Join) -> str:
844            if (
845                expression.side == "LEFT"
846                and not expression.args.get("on")
847                and isinstance(expression.this, exp.Unnest)
848            ):
849                # Some dialects support `LEFT JOIN UNNEST(...)` without an explicit ON clause
850                # DuckDB doesn't, but we can just add a dummy ON clause that is always true
851                return super().join_sql(expression.on(exp.true()))
852
853            return super().join_sql(expression)
def generateseries_sql(self, expression: sqlglot.expressions.GenerateSeries) -> str:
855        def generateseries_sql(self, expression: exp.GenerateSeries) -> str:
856            # GENERATE_SERIES(a, b) -> [a, b], RANGE(a, b) -> [a, b)
857            if expression.args.get("is_end_exclusive"):
858                return rename_func("RANGE")(self, expression)
859
860            return self.function_fallback_sql(expression)
def bracket_sql(self, expression: sqlglot.expressions.Bracket) -> str:
862        def bracket_sql(self, expression: exp.Bracket) -> str:
863            this = expression.this
864            if isinstance(this, exp.Array):
865                this.replace(exp.paren(this))
866
867            bracket = super().bracket_sql(expression)
868
869            if not expression.args.get("returns_list_for_maps"):
870                if not this.type:
871                    from sqlglot.optimizer.annotate_types import annotate_types
872
873                    this = annotate_types(this)
874
875                if this.is_type(exp.DataType.Type.MAP):
876                    bracket = f"({bracket})[1]"
877
878            return bracket
def withingroup_sql(self, expression: sqlglot.expressions.WithinGroup) -> str:
880        def withingroup_sql(self, expression: exp.WithinGroup) -> str:
881            expression_sql = self.sql(expression, "expression")
882
883            func = expression.this
884            if isinstance(func, exp.PERCENTILES):
885                # Make the order key the first arg and slide the fraction to the right
886                # https://duckdb.org/docs/sql/aggregates#ordered-set-aggregate-functions
887                order_col = expression.find(exp.Ordered)
888                if order_col:
889                    func.set("expression", func.this)
890                    func.set("this", order_col.this)
891
892            this = self.sql(expression, "this").rstrip(")")
893
894            return f"{this}{expression_sql})"
def length_sql(self, expression: sqlglot.expressions.Length) -> str:
896        def length_sql(self, expression: exp.Length) -> str:
897            arg = expression.this
898
899            # Dialects like BQ and Snowflake also accept binary values as args, so
900            # DDB will attempt to infer the type or resort to case/when resolution
901            if not expression.args.get("binary") or arg.is_string:
902                return self.func("LENGTH", arg)
903
904            if not arg.type:
905                from sqlglot.optimizer.annotate_types import annotate_types
906
907                arg = annotate_types(arg)
908
909            if arg.is_type(*exp.DataType.TEXT_TYPES):
910                return self.func("LENGTH", arg)
911
912            # We need these casts to make duckdb's static type checker happy
913            blob = exp.cast(arg, exp.DataType.Type.VARBINARY)
914            varchar = exp.cast(arg, exp.DataType.Type.VARCHAR)
915
916            case = (
917                exp.case(self.func("TYPEOF", arg))
918                .when(
919                    "'VARCHAR'", exp.Anonymous(this="LENGTH", expressions=[varchar])
920                )  # anonymous to break length_sql recursion
921                .when("'BLOB'", self.func("OCTET_LENGTH", blob))
922            )
923
924            return self.sql(case)
def objectinsert_sql(self, expression: sqlglot.expressions.ObjectInsert) -> str:
926        def objectinsert_sql(self, expression: exp.ObjectInsert) -> str:
927            this = expression.this
928            key = expression.args.get("key")
929            key_sql = key.name if isinstance(key, exp.Expression) else ""
930            value_sql = self.sql(expression, "value")
931
932            kv_sql = f"{key_sql} := {value_sql}"
933
934            # If the input struct is empty e.g. transpiling OBJECT_INSERT(OBJECT_CONSTRUCT(), key, value) from Snowflake
935            # then we can generate STRUCT_PACK which will build it since STRUCT_INSERT({}, key := value) is not valid DuckDB
936            if isinstance(this, exp.Struct) and not this.expressions:
937                return self.func("STRUCT_PACK", kv_sql)
938
939            return self.func("STRUCT_INSERT", this, kv_sql)
def unnest_sql(self, expression: sqlglot.expressions.Unnest) -> str:
941        def unnest_sql(self, expression: exp.Unnest) -> str:
942            explode_array = expression.args.get("explode_array")
943            if explode_array:
944                # In BigQuery, UNNESTing a nested array leads to explosion of the top-level array & struct
945                # This is transpiled to DDB by transforming "FROM UNNEST(...)" to "FROM (SELECT UNNEST(..., max_depth => 2))"
946                expression.expressions.append(
947                    exp.Kwarg(this=exp.var("max_depth"), expression=exp.Literal.number(2))
948                )
949
950                # If BQ's UNNEST is aliased, we transform it from a column alias to a table alias in DDB
951                alias = expression.args.get("alias")
952                if alias:
953                    expression.set("alias", None)
954                    alias = exp.TableAlias(this=seq_get(alias.args.get("columns"), 0))
955
956                unnest_sql = super().unnest_sql(expression)
957                select = exp.Select(expressions=[unnest_sql]).subquery(alias)
958                return self.sql(select)
959
960            return super().unnest_sql(expression)
def ignorenulls_sql(self, expression: sqlglot.expressions.IgnoreNulls) -> str:
962        def ignorenulls_sql(self, expression: exp.IgnoreNulls) -> str:
963            if isinstance(expression.this, WINDOW_FUNCS_WITH_IGNORE_NULLS):
964                # DuckDB should render IGNORE NULLS only for the general-purpose
965                # window functions that accept it e.g. FIRST_VALUE(... IGNORE NULLS) OVER (...)
966                return super().ignorenulls_sql(expression)
967
968            return self.sql(expression, "this")
def arraytostring_sql(self, expression: sqlglot.expressions.ArrayToString) -> str:
970        def arraytostring_sql(self, expression: exp.ArrayToString) -> str:
971            this = self.sql(expression, "this")
972            null_text = self.sql(expression, "null")
973
974            if null_text:
975                this = f"LIST_TRANSFORM({this}, x -> COALESCE(x, {null_text}))"
976
977            return self.func("ARRAY_TO_STRING", this, expression.expression)
@unsupported_args('position', 'occurrence')
def regexpextract_sql(self, expression: sqlglot.expressions.RegexpExtract) -> str:
979        @unsupported_args("position", "occurrence")
980        def regexpextract_sql(self, expression: exp.RegexpExtract) -> str:
981            group = expression.args.get("group")
982            params = expression.args.get("parameters")
983
984            # Do not render group if there is no following argument,
985            # and it's the default value for this dialect
986            if (
987                not params
988                and group
989                and group.name == str(self.dialect.REGEXP_EXTRACT_DEFAULT_GROUP)
990            ):
991                group = None
992            return self.func(
993                "REGEXP_EXTRACT", expression.this, expression.expression, group, params
994            )
SELECT_KINDS: Tuple[str, ...] = ()
TRY_SUPPORTED = False
SUPPORTS_UESCAPE = False
AFTER_HAVING_MODIFIER_TRANSFORMS = {'windows': <function Generator.<lambda>>, 'qualify': <function Generator.<lambda>>}
Inherited Members
sqlglot.generator.Generator
Generator
NULL_ORDERING_SUPPORTED
LOCKING_READS_SUPPORTED
EXCEPT_INTERSECT_SUPPORT_ALL_CLAUSE
WRAP_DERIVED_VALUES
CREATE_FUNCTION_RETURN_AS
MATCHED_BY_SOURCE
SINGLE_STRING_INTERVAL
INTERVAL_ALLOWS_PLURAL_FORM
LIMIT_ONLY_LITERALS
GROUPINGS_SEP
INDEX_ON
QUERY_HINT_SEP
IS_BOOL_ALLOWED
DUPLICATE_KEY_UPDATE_WITH_SET
LIMIT_IS_TOP
RETURNING_END
EXTRACT_ALLOWS_QUOTES
TZ_TO_WITH_TIME_ZONE
VALUES_AS_TABLE
ALTER_TABLE_INCLUDE_COLUMN_KEYWORD
UNNEST_WITH_ORDINALITY
AGGREGATE_FILTER_SUPPORTED
COMPUTED_COLUMN_WITH_TYPE
SUPPORTS_TABLE_COPY
TABLESAMPLE_REQUIRES_PARENS
TABLESAMPLE_SIZE_IS_ROWS
TABLESAMPLE_WITH_METHOD
COLLATE_IS_FUNC
DATA_TYPE_SPECIFIERS_ALLOWED
ENSURE_BOOLS
CTE_RECURSIVE_KEYWORD_REQUIRED
SUPPORTS_SINGLE_ARG_CONCAT
SUPPORTS_TABLE_ALIAS_COLUMNS
UNPIVOT_ALIASES_ARE_IDENTIFIERS
INSERT_OVERWRITE
SUPPORTS_SELECT_INTO
SUPPORTS_UNLOGGED_TABLES
LIKE_PROPERTY_INSIDE_SCHEMA
JSON_TYPE_REQUIRED_FOR_EXTRACTION
JSON_PATH_SINGLE_QUOTE_ESCAPE
SET_OP_MODIFIERS
COPY_PARAMS_ARE_WRAPPED
COPY_PARAMS_EQ_REQUIRED
HEX_FUNC
WITH_PROPERTIES_PREFIX
QUOTE_JSON_PATH
SUPPORTS_EXPLODING_PROJECTIONS
SUPPORTS_CONVERT_TIMEZONE
SUPPORTS_MEDIAN
SUPPORTS_UNIX_SECONDS
PARSE_JSON_NAME
ARRAY_SIZE_NAME
TIME_PART_SINGULARS
TOKEN_MAPPING
WITH_SEPARATED_COMMENTS
EXCLUDE_COMMENTS
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_parts
column_sql
columnposition_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
set_operation
set_operations
fetch_sql
filter_sql
hint_sql
indexparameters_sql
index_sql
identifier_sql
hex_sql
lowerhex_sql
inputoutputformat_sql
national_sql
partition_sql
properties_sql
root_properties
properties
with_properties
locate_properties
property_name
property_sql
likeproperty_sql
fallbackproperty_sql
journalproperty_sql
freespaceproperty_sql
checksumproperty_sql
mergeblockratioproperty_sql
datablocksizeproperty_sql
blockcompressionproperty_sql
isolatedloadingproperty_sql
partitionboundspec_sql
partitionedofproperty_sql
lockingproperty_sql
withdataproperty_sql
withsystemversioningproperty_sql
insert_sql
introducer_sql
kill_sql
pseudotype_sql
objectidentifier_sql
onconflict_sql
returning_sql
rowformatdelimitedproperty_sql
withtablehint_sql
indextablehint_sql
historicaldata_sql
table_parts
table_sql
pivot_sql
version_sql
tuple_sql
update_sql
values_sql
var_sql
into_sql
from_sql
groupingsets_sql
rollup_sql
cube_sql
group_sql
having_sql
connect_sql
prior_sql
lambda_sql
lateral_op
lateral_sql
limit_sql
offset_sql
setitem_sql
set_sql
pragma_sql
lock_sql
literal_sql
escape_str
loaddata_sql
null_sql
boolean_sql
order_sql
withfill_sql
cluster_sql
distribute_sql
sort_sql
ordered_sql
matchrecognizemeasure_sql
matchrecognize_sql
query_modifiers
options_modifier
queryoption_sql
offset_limit_modifiers
after_limit_modifiers
select_sql
schema_sql
schema_columns_sql
star_sql
parameter_sql
sessionparameter_sql
placeholder_sql
subquery_sql
qualify_sql
prewhere_sql
where_sql
window_sql
partition_by_sql
windowspec_sql
between_sql
bracket_offset_expressions
all_sql
any_sql
exists_sql
case_sql
constraint_sql
nextvaluefor_sql
extract_sql
trim_sql
convert_concat_args
concat_sql
concatws_sql
check_sql
foreignkey_sql
primarykey_sql
if_sql
matchagainst_sql
jsonkeyvalue_sql
jsonpath_sql
json_path_part
formatjson_sql
jsonobject_sql
jsonobjectagg_sql
jsonarray_sql
jsonarrayagg_sql
jsoncolumndef_sql
jsonschema_sql
jsontable_sql
openjsoncolumndef_sql
openjson_sql
in_sql
in_unnest_op
return_sql
reference_sql
anonymous_sql
paren_sql
neg_sql
not_sql
alias_sql
pivotalias_sql
aliases_sql
atindex_sql
attimezone_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
collate_sql
command_sql
comment_sql
mergetreettlaction_sql
mergetreettl_sql
transaction_sql
commit_sql
rollback_sql
altercolumn_sql
alterdiststyle_sql
altersortkey_sql
alterrename_sql
renamecolumn_sql
alterset_sql
alter_sql
add_column_sql
droppartition_sql
addconstraint_sql
distinct_sql
respectnulls_sql
havingmax_sql
intdiv_sql
dpipe_sql
div_sql
overlaps_sql
distance_sql
dot_sql
eq_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
trycast_sql
try_sql
log_sql
use_sql
binary
function_fallback_sql
func
format_args
too_wide
format_time
expressions
op_expressions
naked_property
tag_sql
token_sql
userdefinedfunction_sql
joinhint_sql
kwarg_sql
when_sql
merge_sql
tochar_sql
tonumber_sql
dictproperty_sql
dictrange_sql
dictsubproperty_sql
duplicatekeyproperty_sql
distributedbyproperty_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
toarray_sql
tsordstotime_sql
tsordstotimestamp_sql
tsordstodate_sql
unixdate_sql
lastday_sql
dateadd_sql
arrayany_sql
struct_sql
partitionrange_sql
truncatetable_sql
convert_sql
copyparameter_sql
credentials_sql
copy_sql
semicolon_sql
datadeletionproperty_sql
maskingpolicycolumnconstraint_sql
gapfill_sql
scope_resolution
scoperesolution_sql
rand_sql
changes_sql
pad_sql
summarize_sql
explodinggenerateseries_sql
arrayconcat_sql
converttimezone_sql
json_sql
jsonvalue_sql
conditionalinsert_sql
multitableinserts_sql
oncondition_sql
jsonexists_sql
arrayagg_sql
apply_sql
grant_sql
grantprivilege_sql
grantprincipal_sql
columns_sql
overlay_sql
todouble_sql
string_sql
median_sql
overflowtruncatebehavior_sql
unixseconds_sql
arraysize_sql