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    regexp_extract_sql,
 29    rename_func,
 30    str_position_sql,
 31    str_to_time_sql,
 32    timestamptrunc_sql,
 33    timestrtotime_sql,
 34    unit_to_var,
 35)
 36from sqlglot.helper import seq_get
 37from sqlglot.tokens import TokenType
 38
 39DATETIME_DELTA = t.Union[
 40    exp.DateAdd, exp.TimeAdd, exp.DatetimeAdd, exp.TsOrDsAdd, exp.DateSub, exp.DatetimeSub
 41]
 42
 43
 44def _date_delta_sql(self: DuckDB.Generator, expression: DATETIME_DELTA) -> str:
 45    this = expression.this
 46    unit = unit_to_var(expression)
 47    op = (
 48        "+"
 49        if isinstance(expression, (exp.DateAdd, exp.TimeAdd, exp.DatetimeAdd, exp.TsOrDsAdd))
 50        else "-"
 51    )
 52
 53    to_type: t.Optional[DATA_TYPE] = None
 54    if isinstance(expression, exp.TsOrDsAdd):
 55        to_type = expression.return_type
 56    elif this.is_string:
 57        # Cast string literals (i.e function parameters) to the appropriate type for +/- interval to work
 58        to_type = (
 59            exp.DataType.Type.DATETIME
 60            if isinstance(expression, (exp.DatetimeAdd, exp.DatetimeSub))
 61            else exp.DataType.Type.DATE
 62        )
 63
 64    this = exp.cast(this, to_type) if to_type else this
 65
 66    return f"{self.sql(this)} {op} {self.sql(exp.Interval(this=expression.expression, unit=unit))}"
 67
 68
 69# BigQuery -> DuckDB conversion for the DATE function
 70def _date_sql(self: DuckDB.Generator, expression: exp.Date) -> str:
 71    result = f"CAST({self.sql(expression, 'this')} AS DATE)"
 72    zone = self.sql(expression, "zone")
 73
 74    if zone:
 75        date_str = self.func("STRFTIME", result, "'%d/%m/%Y'")
 76        date_str = f"{date_str} || ' ' || {zone}"
 77
 78        # This will create a TIMESTAMP with time zone information
 79        result = self.func("STRPTIME", date_str, "'%d/%m/%Y %Z'")
 80
 81    return result
 82
 83
 84def _array_sort_sql(self: DuckDB.Generator, expression: exp.ArraySort) -> str:
 85    if expression.expression:
 86        self.unsupported("DuckDB ARRAY_SORT does not support a comparator")
 87    return self.func("ARRAY_SORT", expression.this)
 88
 89
 90def _sort_array_sql(self: DuckDB.Generator, expression: exp.SortArray) -> str:
 91    name = "ARRAY_REVERSE_SORT" if expression.args.get("asc") == exp.false() else "ARRAY_SORT"
 92    return self.func(name, expression.this)
 93
 94
 95def _build_sort_array_desc(args: t.List) -> exp.Expression:
 96    return exp.SortArray(this=seq_get(args, 0), asc=exp.false())
 97
 98
 99def _build_date_diff(args: t.List) -> exp.Expression:
100    return exp.DateDiff(this=seq_get(args, 2), expression=seq_get(args, 1), unit=seq_get(args, 0))
101
102
103def _build_generate_series(end_exclusive: bool = False) -> t.Callable[[t.List], exp.GenerateSeries]:
104    def _builder(args: t.List) -> exp.GenerateSeries:
105        # Check https://duckdb.org/docs/sql/functions/nested.html#range-functions
106        if len(args) == 1:
107            # DuckDB uses 0 as a default for the series' start when it's omitted
108            args.insert(0, exp.Literal.number("0"))
109
110        gen_series = exp.GenerateSeries.from_arg_list(args)
111        gen_series.set("is_end_exclusive", end_exclusive)
112
113        return gen_series
114
115    return _builder
116
117
118def _build_make_timestamp(args: t.List) -> exp.Expression:
119    if len(args) == 1:
120        return exp.UnixToTime(this=seq_get(args, 0), scale=exp.UnixToTime.MICROS)
121
122    return exp.TimestampFromParts(
123        year=seq_get(args, 0),
124        month=seq_get(args, 1),
125        day=seq_get(args, 2),
126        hour=seq_get(args, 3),
127        min=seq_get(args, 4),
128        sec=seq_get(args, 5),
129    )
130
131
132def _struct_sql(self: DuckDB.Generator, expression: exp.Struct) -> str:
133    args: t.List[str] = []
134
135    # BigQuery allows inline construction such as "STRUCT<a STRING, b INTEGER>('str', 1)" which is
136    # canonicalized to "ROW('str', 1) AS STRUCT(a TEXT, b INT)" in DuckDB
137    # The transformation to ROW will take place if a cast to STRUCT / ARRAY of STRUCTs is found
138    ancestor_cast = expression.find_ancestor(exp.Cast)
139    is_struct_cast = ancestor_cast and any(
140        casted_type.is_type(exp.DataType.Type.STRUCT)
141        for casted_type in ancestor_cast.find_all(exp.DataType)
142    )
143
144    for i, expr in enumerate(expression.expressions):
145        is_property_eq = isinstance(expr, exp.PropertyEQ)
146        value = expr.expression if is_property_eq else expr
147
148        if is_struct_cast:
149            args.append(self.sql(value))
150        else:
151            key = expr.name if is_property_eq else f"_{i}"
152            args.append(f"{self.sql(exp.Literal.string(key))}: {self.sql(value)}")
153
154    csv_args = ", ".join(args)
155
156    return f"ROW({csv_args})" if is_struct_cast else f"{{{csv_args}}}"
157
158
159def _datatype_sql(self: DuckDB.Generator, expression: exp.DataType) -> str:
160    if expression.is_type("array"):
161        return f"{self.expressions(expression, flat=True)}[{self.expressions(expression, key='values', flat=True)}]"
162
163    # Type TIMESTAMP / TIME WITH TIME ZONE does not support any modifiers
164    if expression.is_type("timestamptz", "timetz"):
165        return expression.this.value
166
167    return self.datatype_sql(expression)
168
169
170def _json_format_sql(self: DuckDB.Generator, expression: exp.JSONFormat) -> str:
171    sql = self.func("TO_JSON", expression.this, expression.args.get("options"))
172    return f"CAST({sql} AS TEXT)"
173
174
175def _unix_to_time_sql(self: DuckDB.Generator, expression: exp.UnixToTime) -> str:
176    scale = expression.args.get("scale")
177    timestamp = expression.this
178
179    if scale in (None, exp.UnixToTime.SECONDS):
180        return self.func("TO_TIMESTAMP", timestamp)
181    if scale == exp.UnixToTime.MILLIS:
182        return self.func("EPOCH_MS", timestamp)
183    if scale == exp.UnixToTime.MICROS:
184        return self.func("MAKE_TIMESTAMP", timestamp)
185
186    return self.func("TO_TIMESTAMP", exp.Div(this=timestamp, expression=exp.func("POW", 10, scale)))
187
188
189WRAPPED_JSON_EXTRACT_EXPRESSIONS = (exp.Binary, exp.Bracket, exp.In)
190
191
192def _arrow_json_extract_sql(self: DuckDB.Generator, expression: JSON_EXTRACT_TYPE) -> str:
193    arrow_sql = arrow_json_extract_sql(self, expression)
194    if not expression.same_parent and isinstance(
195        expression.parent, WRAPPED_JSON_EXTRACT_EXPRESSIONS
196    ):
197        arrow_sql = self.wrap(arrow_sql)
198    return arrow_sql
199
200
201class DuckDB(Dialect):
202    NULL_ORDERING = "nulls_are_last"
203    SUPPORTS_USER_DEFINED_TYPES = False
204    SAFE_DIVISION = True
205    INDEX_OFFSET = 1
206    CONCAT_COALESCE = True
207    SUPPORTS_ORDER_BY_ALL = True
208
209    # https://duckdb.org/docs/sql/introduction.html#creating-a-new-table
210    NORMALIZATION_STRATEGY = NormalizationStrategy.CASE_INSENSITIVE
211
212    def to_json_path(self, path: t.Optional[exp.Expression]) -> t.Optional[exp.Expression]:
213        if isinstance(path, exp.Literal):
214            # DuckDB also supports the JSON pointer syntax, where every path starts with a `/`.
215            # Additionally, it allows accessing the back of lists using the `[#-i]` syntax.
216            # This check ensures we'll avoid trying to parse these as JSON paths, which can
217            # either result in a noisy warning or in an invalid representation of the path.
218            path_text = path.name
219            if path_text.startswith("/") or "[#" in path_text:
220                return path
221
222        return super().to_json_path(path)
223
224    class Tokenizer(tokens.Tokenizer):
225        HEREDOC_STRINGS = ["$"]
226
227        HEREDOC_TAG_IS_IDENTIFIER = True
228        HEREDOC_STRING_ALTERNATIVE = TokenType.PARAMETER
229
230        KEYWORDS = {
231            **tokens.Tokenizer.KEYWORDS,
232            "//": TokenType.DIV,
233            "ATTACH": TokenType.COMMAND,
234            "BINARY": TokenType.VARBINARY,
235            "BITSTRING": TokenType.BIT,
236            "BPCHAR": TokenType.TEXT,
237            "CHAR": TokenType.TEXT,
238            "CHARACTER VARYING": TokenType.TEXT,
239            "EXCLUDE": TokenType.EXCEPT,
240            "LOGICAL": TokenType.BOOLEAN,
241            "ONLY": TokenType.ONLY,
242            "PIVOT_WIDER": TokenType.PIVOT,
243            "POSITIONAL": TokenType.POSITIONAL,
244            "SIGNED": TokenType.INT,
245            "STRING": TokenType.TEXT,
246            "SUMMARIZE": TokenType.SUMMARIZE,
247            "TIMESTAMP_S": TokenType.TIMESTAMP_S,
248            "TIMESTAMP_MS": TokenType.TIMESTAMP_MS,
249            "TIMESTAMP_NS": TokenType.TIMESTAMP_NS,
250            "TIMESTAMP_US": TokenType.TIMESTAMP,
251            "UBIGINT": TokenType.UBIGINT,
252            "UINTEGER": TokenType.UINT,
253            "USMALLINT": TokenType.USMALLINT,
254            "UTINYINT": TokenType.UTINYINT,
255            "VARCHAR": TokenType.TEXT,
256        }
257        KEYWORDS.pop("/*+")
258
259        SINGLE_TOKENS = {
260            **tokens.Tokenizer.SINGLE_TOKENS,
261            "$": TokenType.PARAMETER,
262        }
263
264    class Parser(parser.Parser):
265        BITWISE = {
266            **parser.Parser.BITWISE,
267            TokenType.TILDA: exp.RegexpLike,
268        }
269
270        FUNCTIONS_WITH_ALIASED_ARGS = {*parser.Parser.FUNCTIONS_WITH_ALIASED_ARGS, "STRUCT_PACK"}
271
272        FUNCTIONS = {
273            **parser.Parser.FUNCTIONS,
274            "ARRAY_REVERSE_SORT": _build_sort_array_desc,
275            "ARRAY_SORT": exp.SortArray.from_arg_list,
276            "DATEDIFF": _build_date_diff,
277            "DATE_DIFF": _build_date_diff,
278            "DATE_TRUNC": date_trunc_to_time,
279            "DATETRUNC": date_trunc_to_time,
280            "DECODE": lambda args: exp.Decode(
281                this=seq_get(args, 0), charset=exp.Literal.string("utf-8")
282            ),
283            "ENCODE": lambda args: exp.Encode(
284                this=seq_get(args, 0), charset=exp.Literal.string("utf-8")
285            ),
286            "EPOCH": exp.TimeToUnix.from_arg_list,
287            "EPOCH_MS": lambda args: exp.UnixToTime(
288                this=seq_get(args, 0), scale=exp.UnixToTime.MILLIS
289            ),
290            "JSON": exp.ParseJSON.from_arg_list,
291            "JSON_EXTRACT_PATH": parser.build_extract_json_with_path(exp.JSONExtract),
292            "JSON_EXTRACT_STRING": parser.build_extract_json_with_path(exp.JSONExtractScalar),
293            "LIST_HAS": exp.ArrayContains.from_arg_list,
294            "LIST_REVERSE_SORT": _build_sort_array_desc,
295            "LIST_SORT": exp.SortArray.from_arg_list,
296            "LIST_VALUE": exp.Array.from_arg_list,
297            "MAKE_TIME": exp.TimeFromParts.from_arg_list,
298            "MAKE_TIMESTAMP": _build_make_timestamp,
299            "MEDIAN": lambda args: exp.PercentileCont(
300                this=seq_get(args, 0), expression=exp.Literal.number(0.5)
301            ),
302            "QUANTILE_CONT": exp.PercentileCont.from_arg_list,
303            "QUANTILE_DISC": exp.PercentileDisc.from_arg_list,
304            "REGEXP_EXTRACT": lambda args: exp.RegexpExtract(
305                this=seq_get(args, 0), expression=seq_get(args, 1), group=seq_get(args, 2)
306            ),
307            "REGEXP_MATCHES": exp.RegexpLike.from_arg_list,
308            "REGEXP_REPLACE": lambda args: exp.RegexpReplace(
309                this=seq_get(args, 0),
310                expression=seq_get(args, 1),
311                replacement=seq_get(args, 2),
312                modifiers=seq_get(args, 3),
313            ),
314            "STRFTIME": build_formatted_time(exp.TimeToStr, "duckdb"),
315            "STRING_SPLIT": exp.Split.from_arg_list,
316            "STRING_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
317            "STRING_TO_ARRAY": exp.Split.from_arg_list,
318            "STRPTIME": build_formatted_time(exp.StrToTime, "duckdb"),
319            "STRUCT_PACK": exp.Struct.from_arg_list,
320            "STR_SPLIT": exp.Split.from_arg_list,
321            "STR_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
322            "TO_TIMESTAMP": exp.UnixToTime.from_arg_list,
323            "UNNEST": exp.Explode.from_arg_list,
324            "XOR": binary_from_function(exp.BitwiseXor),
325            "GENERATE_SERIES": _build_generate_series(),
326            "RANGE": _build_generate_series(end_exclusive=True),
327        }
328
329        FUNCTIONS.pop("DATE_SUB")
330
331        FUNCTION_PARSERS = parser.Parser.FUNCTION_PARSERS.copy()
332        FUNCTION_PARSERS.pop("DECODE")
333
334        NO_PAREN_FUNCTION_PARSERS = {
335            **parser.Parser.NO_PAREN_FUNCTION_PARSERS,
336            "MAP": lambda self: self._parse_map(),
337        }
338
339        TABLE_ALIAS_TOKENS = parser.Parser.TABLE_ALIAS_TOKENS - {
340            TokenType.SEMI,
341            TokenType.ANTI,
342        }
343
344        PLACEHOLDER_PARSERS = {
345            **parser.Parser.PLACEHOLDER_PARSERS,
346            TokenType.PARAMETER: lambda self: (
347                self.expression(exp.Placeholder, this=self._prev.text)
348                if self._match(TokenType.NUMBER) or self._match_set(self.ID_VAR_TOKENS)
349                else None
350            ),
351        }
352
353        TYPE_CONVERTERS = {
354            # https://duckdb.org/docs/sql/data_types/numeric
355            exp.DataType.Type.DECIMAL: build_default_decimal_type(precision=18, scale=3),
356            # https://duckdb.org/docs/sql/data_types/text
357            exp.DataType.Type.TEXT: lambda dtype: exp.DataType.build("TEXT"),
358        }
359
360        def _parse_table_sample(self, as_modifier: bool = False) -> t.Optional[exp.TableSample]:
361            # https://duckdb.org/docs/sql/samples.html
362            sample = super()._parse_table_sample(as_modifier=as_modifier)
363            if sample and not sample.args.get("method"):
364                if sample.args.get("size"):
365                    sample.set("method", exp.var("RESERVOIR"))
366                else:
367                    sample.set("method", exp.var("SYSTEM"))
368
369            return sample
370
371        def _parse_bracket(
372            self, this: t.Optional[exp.Expression] = None
373        ) -> t.Optional[exp.Expression]:
374            bracket = super()._parse_bracket(this)
375            if isinstance(bracket, exp.Bracket):
376                bracket.set("returns_list_for_maps", True)
377
378            return bracket
379
380        def _parse_map(self) -> exp.ToMap | exp.Map:
381            if self._match(TokenType.L_BRACE, advance=False):
382                return self.expression(exp.ToMap, this=self._parse_bracket())
383
384            args = self._parse_wrapped_csv(self._parse_assignment)
385            return self.expression(exp.Map, keys=seq_get(args, 0), values=seq_get(args, 1))
386
387        def _parse_struct_types(self, type_required: bool = False) -> t.Optional[exp.Expression]:
388            return self._parse_field_def()
389
390        def _pivot_column_names(self, aggregations: t.List[exp.Expression]) -> t.List[str]:
391            if len(aggregations) == 1:
392                return super()._pivot_column_names(aggregations)
393            return pivot_column_names(aggregations, dialect="duckdb")
394
395    class Generator(generator.Generator):
396        PARAMETER_TOKEN = "$"
397        NAMED_PLACEHOLDER_TOKEN = "$"
398        JOIN_HINTS = False
399        TABLE_HINTS = False
400        QUERY_HINTS = False
401        LIMIT_FETCH = "LIMIT"
402        STRUCT_DELIMITER = ("(", ")")
403        RENAME_TABLE_WITH_DB = False
404        NVL2_SUPPORTED = False
405        SEMI_ANTI_JOIN_WITH_SIDE = False
406        TABLESAMPLE_KEYWORDS = "USING SAMPLE"
407        TABLESAMPLE_SEED_KEYWORD = "REPEATABLE"
408        LAST_DAY_SUPPORTS_DATE_PART = False
409        JSON_KEY_VALUE_PAIR_SEP = ","
410        IGNORE_NULLS_IN_FUNC = True
411        JSON_PATH_BRACKETED_KEY_SUPPORTED = False
412        SUPPORTS_CREATE_TABLE_LIKE = False
413        MULTI_ARG_DISTINCT = False
414        CAN_IMPLEMENT_ARRAY_ANY = True
415        SUPPORTS_TO_NUMBER = False
416        COPY_HAS_INTO_KEYWORD = False
417        STAR_EXCEPT = "EXCLUDE"
418        PAD_FILL_PATTERN_IS_REQUIRED = True
419
420        TRANSFORMS = {
421            **generator.Generator.TRANSFORMS,
422            exp.ApproxDistinct: approx_count_distinct_sql,
423            exp.Array: inline_array_unless_query,
424            exp.ArrayContainsAll: rename_func("ARRAY_HAS_ALL"),
425            exp.ArrayFilter: rename_func("LIST_FILTER"),
426            exp.ArraySize: rename_func("ARRAY_LENGTH"),
427            exp.ArgMax: arg_max_or_min_no_count("ARG_MAX"),
428            exp.ArgMin: arg_max_or_min_no_count("ARG_MIN"),
429            exp.ArraySort: _array_sort_sql,
430            exp.ArraySum: rename_func("LIST_SUM"),
431            exp.BitwiseXor: rename_func("XOR"),
432            exp.CommentColumnConstraint: no_comment_column_constraint_sql,
433            exp.CurrentDate: lambda *_: "CURRENT_DATE",
434            exp.CurrentTime: lambda *_: "CURRENT_TIME",
435            exp.CurrentTimestamp: lambda *_: "CURRENT_TIMESTAMP",
436            exp.DayOfMonth: rename_func("DAYOFMONTH"),
437            exp.DayOfWeek: rename_func("DAYOFWEEK"),
438            exp.DayOfYear: rename_func("DAYOFYEAR"),
439            exp.DataType: _datatype_sql,
440            exp.Date: _date_sql,
441            exp.DateAdd: _date_delta_sql,
442            exp.DateFromParts: rename_func("MAKE_DATE"),
443            exp.DateSub: _date_delta_sql,
444            exp.DateDiff: lambda self, e: self.func(
445                "DATE_DIFF", f"'{e.args.get('unit') or 'DAY'}'", e.expression, e.this
446            ),
447            exp.DateStrToDate: datestrtodate_sql,
448            exp.Datetime: no_datetime_sql,
449            exp.DatetimeSub: _date_delta_sql,
450            exp.DatetimeAdd: _date_delta_sql,
451            exp.DateToDi: lambda self,
452            e: f"CAST(STRFTIME({self.sql(e, 'this')}, {DuckDB.DATEINT_FORMAT}) AS INT)",
453            exp.Decode: lambda self, e: encode_decode_sql(self, e, "DECODE", replace=False),
454            exp.DiToDate: lambda self,
455            e: f"CAST(STRPTIME(CAST({self.sql(e, 'this')} AS TEXT), {DuckDB.DATEINT_FORMAT}) AS DATE)",
456            exp.Encode: lambda self, e: encode_decode_sql(self, e, "ENCODE", replace=False),
457            exp.Explode: rename_func("UNNEST"),
458            exp.IntDiv: lambda self, e: self.binary(e, "//"),
459            exp.IsInf: rename_func("ISINF"),
460            exp.IsNan: rename_func("ISNAN"),
461            exp.JSONExtract: _arrow_json_extract_sql,
462            exp.JSONExtractScalar: _arrow_json_extract_sql,
463            exp.JSONFormat: _json_format_sql,
464            exp.LogicalOr: rename_func("BOOL_OR"),
465            exp.LogicalAnd: rename_func("BOOL_AND"),
466            exp.MD5Digest: lambda self, e: self.func("UNHEX", self.func("MD5", e.this)),
467            exp.MonthsBetween: lambda self, e: self.func(
468                "DATEDIFF",
469                "'month'",
470                exp.cast(e.expression, exp.DataType.Type.TIMESTAMP, copy=True),
471                exp.cast(e.this, exp.DataType.Type.TIMESTAMP, copy=True),
472            ),
473            exp.PercentileCont: rename_func("QUANTILE_CONT"),
474            exp.PercentileDisc: rename_func("QUANTILE_DISC"),
475            # DuckDB doesn't allow qualified columns inside of PIVOT expressions.
476            # See: https://github.com/duckdb/duckdb/blob/671faf92411182f81dce42ac43de8bfb05d9909e/src/planner/binder/tableref/bind_pivot.cpp#L61-L62
477            exp.Pivot: transforms.preprocess([transforms.unqualify_columns]),
478            exp.RegexpExtract: regexp_extract_sql,
479            exp.RegexpReplace: lambda self, e: self.func(
480                "REGEXP_REPLACE",
481                e.this,
482                e.expression,
483                e.args.get("replacement"),
484                e.args.get("modifiers"),
485            ),
486            exp.RegexpLike: rename_func("REGEXP_MATCHES"),
487            exp.RegexpSplit: rename_func("STR_SPLIT_REGEX"),
488            exp.Return: lambda self, e: self.sql(e, "this"),
489            exp.ReturnsProperty: lambda self, e: "TABLE" if isinstance(e.this, exp.Schema) else "",
490            exp.Rand: rename_func("RANDOM"),
491            exp.SafeDivide: no_safe_divide_sql,
492            exp.Split: rename_func("STR_SPLIT"),
493            exp.SortArray: _sort_array_sql,
494            exp.StrPosition: str_position_sql,
495            exp.StrToUnix: lambda self, e: self.func(
496                "EPOCH", self.func("STRPTIME", e.this, self.format_time(e))
497            ),
498            exp.Struct: _struct_sql,
499            exp.TimeAdd: _date_delta_sql,
500            exp.Time: no_time_sql,
501            exp.Timestamp: no_timestamp_sql,
502            exp.TimestampDiff: lambda self, e: self.func(
503                "DATE_DIFF", exp.Literal.string(e.unit), e.expression, e.this
504            ),
505            exp.TimestampTrunc: timestamptrunc_sql(),
506            exp.TimeStrToDate: lambda self, e: self.sql(exp.cast(e.this, exp.DataType.Type.DATE)),
507            exp.TimeStrToTime: timestrtotime_sql,
508            exp.TimeStrToUnix: lambda self, e: self.func(
509                "EPOCH", exp.cast(e.this, exp.DataType.Type.TIMESTAMP)
510            ),
511            exp.TimeToStr: lambda self, e: self.func("STRFTIME", e.this, self.format_time(e)),
512            exp.TimeToUnix: rename_func("EPOCH"),
513            exp.TsOrDiToDi: lambda self,
514            e: f"CAST(SUBSTR(REPLACE(CAST({self.sql(e, 'this')} AS TEXT), '-', ''), 1, 8) AS INT)",
515            exp.TsOrDsAdd: _date_delta_sql,
516            exp.TsOrDsDiff: lambda self, e: self.func(
517                "DATE_DIFF",
518                f"'{e.args.get('unit') or 'DAY'}'",
519                exp.cast(e.expression, exp.DataType.Type.TIMESTAMP),
520                exp.cast(e.this, exp.DataType.Type.TIMESTAMP),
521            ),
522            exp.UnixToStr: lambda self, e: self.func(
523                "STRFTIME", self.func("TO_TIMESTAMP", e.this), self.format_time(e)
524            ),
525            exp.UnixToTime: _unix_to_time_sql,
526            exp.UnixToTimeStr: lambda self, e: f"CAST(TO_TIMESTAMP({self.sql(e, 'this')}) AS TEXT)",
527            exp.VariancePop: rename_func("VAR_POP"),
528            exp.WeekOfYear: rename_func("WEEKOFYEAR"),
529            exp.Xor: bool_xor_sql,
530        }
531
532        SUPPORTED_JSON_PATH_PARTS = {
533            exp.JSONPathKey,
534            exp.JSONPathRoot,
535            exp.JSONPathSubscript,
536            exp.JSONPathWildcard,
537        }
538
539        TYPE_MAPPING = {
540            **generator.Generator.TYPE_MAPPING,
541            exp.DataType.Type.BINARY: "BLOB",
542            exp.DataType.Type.BPCHAR: "TEXT",
543            exp.DataType.Type.CHAR: "TEXT",
544            exp.DataType.Type.FLOAT: "REAL",
545            exp.DataType.Type.NCHAR: "TEXT",
546            exp.DataType.Type.NVARCHAR: "TEXT",
547            exp.DataType.Type.UINT: "UINTEGER",
548            exp.DataType.Type.VARBINARY: "BLOB",
549            exp.DataType.Type.ROWVERSION: "BLOB",
550            exp.DataType.Type.VARCHAR: "TEXT",
551            exp.DataType.Type.TIMESTAMPNTZ: "TIMESTAMP",
552            exp.DataType.Type.TIMESTAMP_S: "TIMESTAMP_S",
553            exp.DataType.Type.TIMESTAMP_MS: "TIMESTAMP_MS",
554            exp.DataType.Type.TIMESTAMP_NS: "TIMESTAMP_NS",
555        }
556
557        # https://github.com/duckdb/duckdb/blob/ff7f24fd8e3128d94371827523dae85ebaf58713/third_party/libpg_query/grammar/keywords/reserved_keywords.list#L1-L77
558        RESERVED_KEYWORDS = {
559            "array",
560            "analyse",
561            "union",
562            "all",
563            "when",
564            "in_p",
565            "default",
566            "create_p",
567            "window",
568            "asymmetric",
569            "to",
570            "else",
571            "localtime",
572            "from",
573            "end_p",
574            "select",
575            "current_date",
576            "foreign",
577            "with",
578            "grant",
579            "session_user",
580            "or",
581            "except",
582            "references",
583            "fetch",
584            "limit",
585            "group_p",
586            "leading",
587            "into",
588            "collate",
589            "offset",
590            "do",
591            "then",
592            "localtimestamp",
593            "check_p",
594            "lateral_p",
595            "current_role",
596            "where",
597            "asc_p",
598            "placing",
599            "desc_p",
600            "user",
601            "unique",
602            "initially",
603            "column",
604            "both",
605            "some",
606            "as",
607            "any",
608            "only",
609            "deferrable",
610            "null_p",
611            "current_time",
612            "true_p",
613            "table",
614            "case",
615            "trailing",
616            "variadic",
617            "for",
618            "on",
619            "distinct",
620            "false_p",
621            "not",
622            "constraint",
623            "current_timestamp",
624            "returning",
625            "primary",
626            "intersect",
627            "having",
628            "analyze",
629            "current_user",
630            "and",
631            "cast",
632            "symmetric",
633            "using",
634            "order",
635            "current_catalog",
636        }
637
638        UNWRAPPED_INTERVAL_VALUES = (exp.Literal, exp.Paren)
639
640        # DuckDB doesn't generally support CREATE TABLE .. properties
641        # https://duckdb.org/docs/sql/statements/create_table.html
642        PROPERTIES_LOCATION = {
643            prop: exp.Properties.Location.UNSUPPORTED
644            for prop in generator.Generator.PROPERTIES_LOCATION
645        }
646
647        # There are a few exceptions (e.g. temporary tables) which are supported or
648        # can be transpiled to DuckDB, so we explicitly override them accordingly
649        PROPERTIES_LOCATION[exp.LikeProperty] = exp.Properties.Location.POST_SCHEMA
650        PROPERTIES_LOCATION[exp.TemporaryProperty] = exp.Properties.Location.POST_CREATE
651        PROPERTIES_LOCATION[exp.ReturnsProperty] = exp.Properties.Location.POST_ALIAS
652
653        def strtotime_sql(self, expression: exp.StrToTime) -> str:
654            if expression.args.get("safe"):
655                formatted_time = self.format_time(expression)
656                return f"CAST({self.func('TRY_STRPTIME', expression.this, formatted_time)} AS TIMESTAMP)"
657            return str_to_time_sql(self, expression)
658
659        def strtodate_sql(self, expression: exp.StrToDate) -> str:
660            if expression.args.get("safe"):
661                formatted_time = self.format_time(expression)
662                return f"CAST({self.func('TRY_STRPTIME', expression.this, formatted_time)} AS DATE)"
663            return f"CAST({str_to_time_sql(self, expression)} AS DATE)"
664
665        def parsejson_sql(self, expression: exp.ParseJSON) -> str:
666            arg = expression.this
667            if expression.args.get("safe"):
668                return self.sql(exp.case().when(exp.func("json_valid", arg), arg).else_(exp.null()))
669            return self.func("JSON", arg)
670
671        def timefromparts_sql(self, expression: exp.TimeFromParts) -> str:
672            nano = expression.args.get("nano")
673            if nano is not None:
674                expression.set(
675                    "sec", expression.args["sec"] + nano.pop() / exp.Literal.number(1000000000.0)
676                )
677
678            return rename_func("MAKE_TIME")(self, expression)
679
680        def timestampfromparts_sql(self, expression: exp.TimestampFromParts) -> str:
681            sec = expression.args["sec"]
682
683            milli = expression.args.get("milli")
684            if milli is not None:
685                sec += milli.pop() / exp.Literal.number(1000.0)
686
687            nano = expression.args.get("nano")
688            if nano is not None:
689                sec += nano.pop() / exp.Literal.number(1000000000.0)
690
691            if milli or nano:
692                expression.set("sec", sec)
693
694            return rename_func("MAKE_TIMESTAMP")(self, expression)
695
696        def tablesample_sql(
697            self,
698            expression: exp.TableSample,
699            sep: str = " AS ",
700            tablesample_keyword: t.Optional[str] = None,
701        ) -> str:
702            if not isinstance(expression.parent, exp.Select):
703                # This sample clause only applies to a single source, not the entire resulting relation
704                tablesample_keyword = "TABLESAMPLE"
705
706            if expression.args.get("size"):
707                method = expression.args.get("method")
708                if method and method.name.upper() != "RESERVOIR":
709                    self.unsupported(
710                        f"Sampling method {method} is not supported with a discrete sample count, "
711                        "defaulting to reservoir sampling"
712                    )
713                    expression.set("method", exp.var("RESERVOIR"))
714
715            return super().tablesample_sql(
716                expression, sep=sep, tablesample_keyword=tablesample_keyword
717            )
718
719        def interval_sql(self, expression: exp.Interval) -> str:
720            multiplier: t.Optional[int] = None
721            unit = expression.text("unit").lower()
722
723            if unit.startswith("week"):
724                multiplier = 7
725            if unit.startswith("quarter"):
726                multiplier = 90
727
728            if multiplier:
729                return f"({multiplier} * {super().interval_sql(exp.Interval(this=expression.this, unit=exp.var('DAY')))})"
730
731            return super().interval_sql(expression)
732
733        def columndef_sql(self, expression: exp.ColumnDef, sep: str = " ") -> str:
734            if isinstance(expression.parent, exp.UserDefinedFunction):
735                return self.sql(expression, "this")
736            return super().columndef_sql(expression, sep)
737
738        def join_sql(self, expression: exp.Join) -> str:
739            if (
740                expression.side == "LEFT"
741                and not expression.args.get("on")
742                and isinstance(expression.this, exp.Unnest)
743            ):
744                # Some dialects support `LEFT JOIN UNNEST(...)` without an explicit ON clause
745                # DuckDB doesn't, but we can just add a dummy ON clause that is always true
746                return super().join_sql(expression.on(exp.true()))
747
748            return super().join_sql(expression)
749
750        def generateseries_sql(self, expression: exp.GenerateSeries) -> str:
751            # GENERATE_SERIES(a, b) -> [a, b], RANGE(a, b) -> [a, b)
752            if expression.args.get("is_end_exclusive"):
753                return rename_func("RANGE")(self, expression)
754
755            return self.function_fallback_sql(expression)
756
757        def bracket_sql(self, expression: exp.Bracket) -> str:
758            this = expression.this
759            if isinstance(this, exp.Array):
760                this.replace(exp.paren(this))
761
762            bracket = super().bracket_sql(expression)
763
764            if not expression.args.get("returns_list_for_maps"):
765                if not this.type:
766                    from sqlglot.optimizer.annotate_types import annotate_types
767
768                    this = annotate_types(this)
769
770                if this.is_type(exp.DataType.Type.MAP):
771                    bracket = f"({bracket})[1]"
772
773            return bracket
774
775        def withingroup_sql(self, expression: exp.WithinGroup) -> str:
776            expression_sql = self.sql(expression, "expression")
777
778            func = expression.this
779            if isinstance(func, exp.PERCENTILES):
780                # Make the order key the first arg and slide the fraction to the right
781                # https://duckdb.org/docs/sql/aggregates#ordered-set-aggregate-functions
782                order_col = expression.find(exp.Ordered)
783                if order_col:
784                    func.set("expression", func.this)
785                    func.set("this", order_col.this)
786
787            this = self.sql(expression, "this").rstrip(")")
788
789            return f"{this}{expression_sql})"
790
791        def length_sql(self, expression: exp.Length) -> str:
792            arg = expression.this
793
794            # Dialects like BQ and Snowflake also accept binary values as args, so
795            # DDB will attempt to infer the type or resort to case/when resolution
796            if not expression.args.get("binary") or arg.is_string:
797                return self.func("LENGTH", arg)
798
799            if not arg.type:
800                from sqlglot.optimizer.annotate_types import annotate_types
801
802                arg = annotate_types(arg)
803
804            if arg.is_type(*exp.DataType.TEXT_TYPES):
805                return self.func("LENGTH", arg)
806
807            # We need these casts to make duckdb's static type checker happy
808            blob = exp.cast(arg, exp.DataType.Type.VARBINARY)
809            varchar = exp.cast(arg, exp.DataType.Type.VARCHAR)
810
811            case = (
812                exp.case(self.func("TYPEOF", arg))
813                .when(
814                    "'VARCHAR'", exp.Anonymous(this="LENGTH", expressions=[varchar])
815                )  # anonymous to break length_sql recursion
816                .when("'BLOB'", self.func("OCTET_LENGTH", blob))
817            )
818
819            return self.sql(case)
820
821        def objectinsert_sql(self, expression: exp.ObjectInsert) -> str:
822            this = expression.this
823            key = expression.args.get("key")
824            key_sql = key.name if isinstance(key, exp.Expression) else ""
825            value_sql = self.sql(expression, "value")
826
827            kv_sql = f"{key_sql} := {value_sql}"
828
829            # If the input struct is empty e.g. transpiling OBJECT_INSERT(OBJECT_CONSTRUCT(), key, value) from Snowflake
830            # then we can generate STRUCT_PACK which will build it since STRUCT_INSERT({}, key := value) is not valid DuckDB
831            if isinstance(this, exp.Struct) and not this.expressions:
832                return self.func("STRUCT_PACK", kv_sql)
833
834            return self.func("STRUCT_INSERT", this, kv_sql)
WRAPPED_JSON_EXTRACT_EXPRESSIONS = (<class 'sqlglot.expressions.Binary'>, <class 'sqlglot.expressions.Bracket'>, <class 'sqlglot.expressions.In'>)
class DuckDB(sqlglot.dialects.dialect.Dialect):
202class DuckDB(Dialect):
203    NULL_ORDERING = "nulls_are_last"
204    SUPPORTS_USER_DEFINED_TYPES = False
205    SAFE_DIVISION = True
206    INDEX_OFFSET = 1
207    CONCAT_COALESCE = True
208    SUPPORTS_ORDER_BY_ALL = True
209
210    # https://duckdb.org/docs/sql/introduction.html#creating-a-new-table
211    NORMALIZATION_STRATEGY = NormalizationStrategy.CASE_INSENSITIVE
212
213    def to_json_path(self, path: t.Optional[exp.Expression]) -> t.Optional[exp.Expression]:
214        if isinstance(path, exp.Literal):
215            # DuckDB also supports the JSON pointer syntax, where every path starts with a `/`.
216            # Additionally, it allows accessing the back of lists using the `[#-i]` syntax.
217            # This check ensures we'll avoid trying to parse these as JSON paths, which can
218            # either result in a noisy warning or in an invalid representation of the path.
219            path_text = path.name
220            if path_text.startswith("/") or "[#" in path_text:
221                return path
222
223        return super().to_json_path(path)
224
225    class Tokenizer(tokens.Tokenizer):
226        HEREDOC_STRINGS = ["$"]
227
228        HEREDOC_TAG_IS_IDENTIFIER = True
229        HEREDOC_STRING_ALTERNATIVE = TokenType.PARAMETER
230
231        KEYWORDS = {
232            **tokens.Tokenizer.KEYWORDS,
233            "//": TokenType.DIV,
234            "ATTACH": TokenType.COMMAND,
235            "BINARY": TokenType.VARBINARY,
236            "BITSTRING": TokenType.BIT,
237            "BPCHAR": TokenType.TEXT,
238            "CHAR": TokenType.TEXT,
239            "CHARACTER VARYING": TokenType.TEXT,
240            "EXCLUDE": TokenType.EXCEPT,
241            "LOGICAL": TokenType.BOOLEAN,
242            "ONLY": TokenType.ONLY,
243            "PIVOT_WIDER": TokenType.PIVOT,
244            "POSITIONAL": TokenType.POSITIONAL,
245            "SIGNED": TokenType.INT,
246            "STRING": TokenType.TEXT,
247            "SUMMARIZE": TokenType.SUMMARIZE,
248            "TIMESTAMP_S": TokenType.TIMESTAMP_S,
249            "TIMESTAMP_MS": TokenType.TIMESTAMP_MS,
250            "TIMESTAMP_NS": TokenType.TIMESTAMP_NS,
251            "TIMESTAMP_US": TokenType.TIMESTAMP,
252            "UBIGINT": TokenType.UBIGINT,
253            "UINTEGER": TokenType.UINT,
254            "USMALLINT": TokenType.USMALLINT,
255            "UTINYINT": TokenType.UTINYINT,
256            "VARCHAR": TokenType.TEXT,
257        }
258        KEYWORDS.pop("/*+")
259
260        SINGLE_TOKENS = {
261            **tokens.Tokenizer.SINGLE_TOKENS,
262            "$": TokenType.PARAMETER,
263        }
264
265    class Parser(parser.Parser):
266        BITWISE = {
267            **parser.Parser.BITWISE,
268            TokenType.TILDA: exp.RegexpLike,
269        }
270
271        FUNCTIONS_WITH_ALIASED_ARGS = {*parser.Parser.FUNCTIONS_WITH_ALIASED_ARGS, "STRUCT_PACK"}
272
273        FUNCTIONS = {
274            **parser.Parser.FUNCTIONS,
275            "ARRAY_REVERSE_SORT": _build_sort_array_desc,
276            "ARRAY_SORT": exp.SortArray.from_arg_list,
277            "DATEDIFF": _build_date_diff,
278            "DATE_DIFF": _build_date_diff,
279            "DATE_TRUNC": date_trunc_to_time,
280            "DATETRUNC": date_trunc_to_time,
281            "DECODE": lambda args: exp.Decode(
282                this=seq_get(args, 0), charset=exp.Literal.string("utf-8")
283            ),
284            "ENCODE": lambda args: exp.Encode(
285                this=seq_get(args, 0), charset=exp.Literal.string("utf-8")
286            ),
287            "EPOCH": exp.TimeToUnix.from_arg_list,
288            "EPOCH_MS": lambda args: exp.UnixToTime(
289                this=seq_get(args, 0), scale=exp.UnixToTime.MILLIS
290            ),
291            "JSON": exp.ParseJSON.from_arg_list,
292            "JSON_EXTRACT_PATH": parser.build_extract_json_with_path(exp.JSONExtract),
293            "JSON_EXTRACT_STRING": parser.build_extract_json_with_path(exp.JSONExtractScalar),
294            "LIST_HAS": exp.ArrayContains.from_arg_list,
295            "LIST_REVERSE_SORT": _build_sort_array_desc,
296            "LIST_SORT": exp.SortArray.from_arg_list,
297            "LIST_VALUE": exp.Array.from_arg_list,
298            "MAKE_TIME": exp.TimeFromParts.from_arg_list,
299            "MAKE_TIMESTAMP": _build_make_timestamp,
300            "MEDIAN": lambda args: exp.PercentileCont(
301                this=seq_get(args, 0), expression=exp.Literal.number(0.5)
302            ),
303            "QUANTILE_CONT": exp.PercentileCont.from_arg_list,
304            "QUANTILE_DISC": exp.PercentileDisc.from_arg_list,
305            "REGEXP_EXTRACT": lambda args: exp.RegexpExtract(
306                this=seq_get(args, 0), expression=seq_get(args, 1), group=seq_get(args, 2)
307            ),
308            "REGEXP_MATCHES": exp.RegexpLike.from_arg_list,
309            "REGEXP_REPLACE": lambda args: exp.RegexpReplace(
310                this=seq_get(args, 0),
311                expression=seq_get(args, 1),
312                replacement=seq_get(args, 2),
313                modifiers=seq_get(args, 3),
314            ),
315            "STRFTIME": build_formatted_time(exp.TimeToStr, "duckdb"),
316            "STRING_SPLIT": exp.Split.from_arg_list,
317            "STRING_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
318            "STRING_TO_ARRAY": exp.Split.from_arg_list,
319            "STRPTIME": build_formatted_time(exp.StrToTime, "duckdb"),
320            "STRUCT_PACK": exp.Struct.from_arg_list,
321            "STR_SPLIT": exp.Split.from_arg_list,
322            "STR_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
323            "TO_TIMESTAMP": exp.UnixToTime.from_arg_list,
324            "UNNEST": exp.Explode.from_arg_list,
325            "XOR": binary_from_function(exp.BitwiseXor),
326            "GENERATE_SERIES": _build_generate_series(),
327            "RANGE": _build_generate_series(end_exclusive=True),
328        }
329
330        FUNCTIONS.pop("DATE_SUB")
331
332        FUNCTION_PARSERS = parser.Parser.FUNCTION_PARSERS.copy()
333        FUNCTION_PARSERS.pop("DECODE")
334
335        NO_PAREN_FUNCTION_PARSERS = {
336            **parser.Parser.NO_PAREN_FUNCTION_PARSERS,
337            "MAP": lambda self: self._parse_map(),
338        }
339
340        TABLE_ALIAS_TOKENS = parser.Parser.TABLE_ALIAS_TOKENS - {
341            TokenType.SEMI,
342            TokenType.ANTI,
343        }
344
345        PLACEHOLDER_PARSERS = {
346            **parser.Parser.PLACEHOLDER_PARSERS,
347            TokenType.PARAMETER: lambda self: (
348                self.expression(exp.Placeholder, this=self._prev.text)
349                if self._match(TokenType.NUMBER) or self._match_set(self.ID_VAR_TOKENS)
350                else None
351            ),
352        }
353
354        TYPE_CONVERTERS = {
355            # https://duckdb.org/docs/sql/data_types/numeric
356            exp.DataType.Type.DECIMAL: build_default_decimal_type(precision=18, scale=3),
357            # https://duckdb.org/docs/sql/data_types/text
358            exp.DataType.Type.TEXT: lambda dtype: exp.DataType.build("TEXT"),
359        }
360
361        def _parse_table_sample(self, as_modifier: bool = False) -> t.Optional[exp.TableSample]:
362            # https://duckdb.org/docs/sql/samples.html
363            sample = super()._parse_table_sample(as_modifier=as_modifier)
364            if sample and not sample.args.get("method"):
365                if sample.args.get("size"):
366                    sample.set("method", exp.var("RESERVOIR"))
367                else:
368                    sample.set("method", exp.var("SYSTEM"))
369
370            return sample
371
372        def _parse_bracket(
373            self, this: t.Optional[exp.Expression] = None
374        ) -> t.Optional[exp.Expression]:
375            bracket = super()._parse_bracket(this)
376            if isinstance(bracket, exp.Bracket):
377                bracket.set("returns_list_for_maps", True)
378
379            return bracket
380
381        def _parse_map(self) -> exp.ToMap | exp.Map:
382            if self._match(TokenType.L_BRACE, advance=False):
383                return self.expression(exp.ToMap, this=self._parse_bracket())
384
385            args = self._parse_wrapped_csv(self._parse_assignment)
386            return self.expression(exp.Map, keys=seq_get(args, 0), values=seq_get(args, 1))
387
388        def _parse_struct_types(self, type_required: bool = False) -> t.Optional[exp.Expression]:
389            return self._parse_field_def()
390
391        def _pivot_column_names(self, aggregations: t.List[exp.Expression]) -> t.List[str]:
392            if len(aggregations) == 1:
393                return super()._pivot_column_names(aggregations)
394            return pivot_column_names(aggregations, dialect="duckdb")
395
396    class Generator(generator.Generator):
397        PARAMETER_TOKEN = "$"
398        NAMED_PLACEHOLDER_TOKEN = "$"
399        JOIN_HINTS = False
400        TABLE_HINTS = False
401        QUERY_HINTS = False
402        LIMIT_FETCH = "LIMIT"
403        STRUCT_DELIMITER = ("(", ")")
404        RENAME_TABLE_WITH_DB = False
405        NVL2_SUPPORTED = False
406        SEMI_ANTI_JOIN_WITH_SIDE = False
407        TABLESAMPLE_KEYWORDS = "USING SAMPLE"
408        TABLESAMPLE_SEED_KEYWORD = "REPEATABLE"
409        LAST_DAY_SUPPORTS_DATE_PART = False
410        JSON_KEY_VALUE_PAIR_SEP = ","
411        IGNORE_NULLS_IN_FUNC = True
412        JSON_PATH_BRACKETED_KEY_SUPPORTED = False
413        SUPPORTS_CREATE_TABLE_LIKE = False
414        MULTI_ARG_DISTINCT = False
415        CAN_IMPLEMENT_ARRAY_ANY = True
416        SUPPORTS_TO_NUMBER = False
417        COPY_HAS_INTO_KEYWORD = False
418        STAR_EXCEPT = "EXCLUDE"
419        PAD_FILL_PATTERN_IS_REQUIRED = True
420
421        TRANSFORMS = {
422            **generator.Generator.TRANSFORMS,
423            exp.ApproxDistinct: approx_count_distinct_sql,
424            exp.Array: inline_array_unless_query,
425            exp.ArrayContainsAll: rename_func("ARRAY_HAS_ALL"),
426            exp.ArrayFilter: rename_func("LIST_FILTER"),
427            exp.ArraySize: rename_func("ARRAY_LENGTH"),
428            exp.ArgMax: arg_max_or_min_no_count("ARG_MAX"),
429            exp.ArgMin: arg_max_or_min_no_count("ARG_MIN"),
430            exp.ArraySort: _array_sort_sql,
431            exp.ArraySum: rename_func("LIST_SUM"),
432            exp.BitwiseXor: rename_func("XOR"),
433            exp.CommentColumnConstraint: no_comment_column_constraint_sql,
434            exp.CurrentDate: lambda *_: "CURRENT_DATE",
435            exp.CurrentTime: lambda *_: "CURRENT_TIME",
436            exp.CurrentTimestamp: lambda *_: "CURRENT_TIMESTAMP",
437            exp.DayOfMonth: rename_func("DAYOFMONTH"),
438            exp.DayOfWeek: rename_func("DAYOFWEEK"),
439            exp.DayOfYear: rename_func("DAYOFYEAR"),
440            exp.DataType: _datatype_sql,
441            exp.Date: _date_sql,
442            exp.DateAdd: _date_delta_sql,
443            exp.DateFromParts: rename_func("MAKE_DATE"),
444            exp.DateSub: _date_delta_sql,
445            exp.DateDiff: lambda self, e: self.func(
446                "DATE_DIFF", f"'{e.args.get('unit') or 'DAY'}'", e.expression, e.this
447            ),
448            exp.DateStrToDate: datestrtodate_sql,
449            exp.Datetime: no_datetime_sql,
450            exp.DatetimeSub: _date_delta_sql,
451            exp.DatetimeAdd: _date_delta_sql,
452            exp.DateToDi: lambda self,
453            e: f"CAST(STRFTIME({self.sql(e, 'this')}, {DuckDB.DATEINT_FORMAT}) AS INT)",
454            exp.Decode: lambda self, e: encode_decode_sql(self, e, "DECODE", replace=False),
455            exp.DiToDate: lambda self,
456            e: f"CAST(STRPTIME(CAST({self.sql(e, 'this')} AS TEXT), {DuckDB.DATEINT_FORMAT}) AS DATE)",
457            exp.Encode: lambda self, e: encode_decode_sql(self, e, "ENCODE", replace=False),
458            exp.Explode: rename_func("UNNEST"),
459            exp.IntDiv: lambda self, e: self.binary(e, "//"),
460            exp.IsInf: rename_func("ISINF"),
461            exp.IsNan: rename_func("ISNAN"),
462            exp.JSONExtract: _arrow_json_extract_sql,
463            exp.JSONExtractScalar: _arrow_json_extract_sql,
464            exp.JSONFormat: _json_format_sql,
465            exp.LogicalOr: rename_func("BOOL_OR"),
466            exp.LogicalAnd: rename_func("BOOL_AND"),
467            exp.MD5Digest: lambda self, e: self.func("UNHEX", self.func("MD5", e.this)),
468            exp.MonthsBetween: lambda self, e: self.func(
469                "DATEDIFF",
470                "'month'",
471                exp.cast(e.expression, exp.DataType.Type.TIMESTAMP, copy=True),
472                exp.cast(e.this, exp.DataType.Type.TIMESTAMP, copy=True),
473            ),
474            exp.PercentileCont: rename_func("QUANTILE_CONT"),
475            exp.PercentileDisc: rename_func("QUANTILE_DISC"),
476            # DuckDB doesn't allow qualified columns inside of PIVOT expressions.
477            # See: https://github.com/duckdb/duckdb/blob/671faf92411182f81dce42ac43de8bfb05d9909e/src/planner/binder/tableref/bind_pivot.cpp#L61-L62
478            exp.Pivot: transforms.preprocess([transforms.unqualify_columns]),
479            exp.RegexpExtract: regexp_extract_sql,
480            exp.RegexpReplace: lambda self, e: self.func(
481                "REGEXP_REPLACE",
482                e.this,
483                e.expression,
484                e.args.get("replacement"),
485                e.args.get("modifiers"),
486            ),
487            exp.RegexpLike: rename_func("REGEXP_MATCHES"),
488            exp.RegexpSplit: rename_func("STR_SPLIT_REGEX"),
489            exp.Return: lambda self, e: self.sql(e, "this"),
490            exp.ReturnsProperty: lambda self, e: "TABLE" if isinstance(e.this, exp.Schema) else "",
491            exp.Rand: rename_func("RANDOM"),
492            exp.SafeDivide: no_safe_divide_sql,
493            exp.Split: rename_func("STR_SPLIT"),
494            exp.SortArray: _sort_array_sql,
495            exp.StrPosition: str_position_sql,
496            exp.StrToUnix: lambda self, e: self.func(
497                "EPOCH", self.func("STRPTIME", e.this, self.format_time(e))
498            ),
499            exp.Struct: _struct_sql,
500            exp.TimeAdd: _date_delta_sql,
501            exp.Time: no_time_sql,
502            exp.Timestamp: no_timestamp_sql,
503            exp.TimestampDiff: lambda self, e: self.func(
504                "DATE_DIFF", exp.Literal.string(e.unit), e.expression, e.this
505            ),
506            exp.TimestampTrunc: timestamptrunc_sql(),
507            exp.TimeStrToDate: lambda self, e: self.sql(exp.cast(e.this, exp.DataType.Type.DATE)),
508            exp.TimeStrToTime: timestrtotime_sql,
509            exp.TimeStrToUnix: lambda self, e: self.func(
510                "EPOCH", exp.cast(e.this, exp.DataType.Type.TIMESTAMP)
511            ),
512            exp.TimeToStr: lambda self, e: self.func("STRFTIME", e.this, self.format_time(e)),
513            exp.TimeToUnix: rename_func("EPOCH"),
514            exp.TsOrDiToDi: lambda self,
515            e: f"CAST(SUBSTR(REPLACE(CAST({self.sql(e, 'this')} AS TEXT), '-', ''), 1, 8) AS INT)",
516            exp.TsOrDsAdd: _date_delta_sql,
517            exp.TsOrDsDiff: lambda self, e: self.func(
518                "DATE_DIFF",
519                f"'{e.args.get('unit') or 'DAY'}'",
520                exp.cast(e.expression, exp.DataType.Type.TIMESTAMP),
521                exp.cast(e.this, exp.DataType.Type.TIMESTAMP),
522            ),
523            exp.UnixToStr: lambda self, e: self.func(
524                "STRFTIME", self.func("TO_TIMESTAMP", e.this), self.format_time(e)
525            ),
526            exp.UnixToTime: _unix_to_time_sql,
527            exp.UnixToTimeStr: lambda self, e: f"CAST(TO_TIMESTAMP({self.sql(e, 'this')}) AS TEXT)",
528            exp.VariancePop: rename_func("VAR_POP"),
529            exp.WeekOfYear: rename_func("WEEKOFYEAR"),
530            exp.Xor: bool_xor_sql,
531        }
532
533        SUPPORTED_JSON_PATH_PARTS = {
534            exp.JSONPathKey,
535            exp.JSONPathRoot,
536            exp.JSONPathSubscript,
537            exp.JSONPathWildcard,
538        }
539
540        TYPE_MAPPING = {
541            **generator.Generator.TYPE_MAPPING,
542            exp.DataType.Type.BINARY: "BLOB",
543            exp.DataType.Type.BPCHAR: "TEXT",
544            exp.DataType.Type.CHAR: "TEXT",
545            exp.DataType.Type.FLOAT: "REAL",
546            exp.DataType.Type.NCHAR: "TEXT",
547            exp.DataType.Type.NVARCHAR: "TEXT",
548            exp.DataType.Type.UINT: "UINTEGER",
549            exp.DataType.Type.VARBINARY: "BLOB",
550            exp.DataType.Type.ROWVERSION: "BLOB",
551            exp.DataType.Type.VARCHAR: "TEXT",
552            exp.DataType.Type.TIMESTAMPNTZ: "TIMESTAMP",
553            exp.DataType.Type.TIMESTAMP_S: "TIMESTAMP_S",
554            exp.DataType.Type.TIMESTAMP_MS: "TIMESTAMP_MS",
555            exp.DataType.Type.TIMESTAMP_NS: "TIMESTAMP_NS",
556        }
557
558        # https://github.com/duckdb/duckdb/blob/ff7f24fd8e3128d94371827523dae85ebaf58713/third_party/libpg_query/grammar/keywords/reserved_keywords.list#L1-L77
559        RESERVED_KEYWORDS = {
560            "array",
561            "analyse",
562            "union",
563            "all",
564            "when",
565            "in_p",
566            "default",
567            "create_p",
568            "window",
569            "asymmetric",
570            "to",
571            "else",
572            "localtime",
573            "from",
574            "end_p",
575            "select",
576            "current_date",
577            "foreign",
578            "with",
579            "grant",
580            "session_user",
581            "or",
582            "except",
583            "references",
584            "fetch",
585            "limit",
586            "group_p",
587            "leading",
588            "into",
589            "collate",
590            "offset",
591            "do",
592            "then",
593            "localtimestamp",
594            "check_p",
595            "lateral_p",
596            "current_role",
597            "where",
598            "asc_p",
599            "placing",
600            "desc_p",
601            "user",
602            "unique",
603            "initially",
604            "column",
605            "both",
606            "some",
607            "as",
608            "any",
609            "only",
610            "deferrable",
611            "null_p",
612            "current_time",
613            "true_p",
614            "table",
615            "case",
616            "trailing",
617            "variadic",
618            "for",
619            "on",
620            "distinct",
621            "false_p",
622            "not",
623            "constraint",
624            "current_timestamp",
625            "returning",
626            "primary",
627            "intersect",
628            "having",
629            "analyze",
630            "current_user",
631            "and",
632            "cast",
633            "symmetric",
634            "using",
635            "order",
636            "current_catalog",
637        }
638
639        UNWRAPPED_INTERVAL_VALUES = (exp.Literal, exp.Paren)
640
641        # DuckDB doesn't generally support CREATE TABLE .. properties
642        # https://duckdb.org/docs/sql/statements/create_table.html
643        PROPERTIES_LOCATION = {
644            prop: exp.Properties.Location.UNSUPPORTED
645            for prop in generator.Generator.PROPERTIES_LOCATION
646        }
647
648        # There are a few exceptions (e.g. temporary tables) which are supported or
649        # can be transpiled to DuckDB, so we explicitly override them accordingly
650        PROPERTIES_LOCATION[exp.LikeProperty] = exp.Properties.Location.POST_SCHEMA
651        PROPERTIES_LOCATION[exp.TemporaryProperty] = exp.Properties.Location.POST_CREATE
652        PROPERTIES_LOCATION[exp.ReturnsProperty] = exp.Properties.Location.POST_ALIAS
653
654        def strtotime_sql(self, expression: exp.StrToTime) -> str:
655            if expression.args.get("safe"):
656                formatted_time = self.format_time(expression)
657                return f"CAST({self.func('TRY_STRPTIME', expression.this, formatted_time)} AS TIMESTAMP)"
658            return str_to_time_sql(self, expression)
659
660        def strtodate_sql(self, expression: exp.StrToDate) -> str:
661            if expression.args.get("safe"):
662                formatted_time = self.format_time(expression)
663                return f"CAST({self.func('TRY_STRPTIME', expression.this, formatted_time)} AS DATE)"
664            return f"CAST({str_to_time_sql(self, expression)} AS DATE)"
665
666        def parsejson_sql(self, expression: exp.ParseJSON) -> str:
667            arg = expression.this
668            if expression.args.get("safe"):
669                return self.sql(exp.case().when(exp.func("json_valid", arg), arg).else_(exp.null()))
670            return self.func("JSON", arg)
671
672        def timefromparts_sql(self, expression: exp.TimeFromParts) -> str:
673            nano = expression.args.get("nano")
674            if nano is not None:
675                expression.set(
676                    "sec", expression.args["sec"] + nano.pop() / exp.Literal.number(1000000000.0)
677                )
678
679            return rename_func("MAKE_TIME")(self, expression)
680
681        def timestampfromparts_sql(self, expression: exp.TimestampFromParts) -> str:
682            sec = expression.args["sec"]
683
684            milli = expression.args.get("milli")
685            if milli is not None:
686                sec += milli.pop() / exp.Literal.number(1000.0)
687
688            nano = expression.args.get("nano")
689            if nano is not None:
690                sec += nano.pop() / exp.Literal.number(1000000000.0)
691
692            if milli or nano:
693                expression.set("sec", sec)
694
695            return rename_func("MAKE_TIMESTAMP")(self, expression)
696
697        def tablesample_sql(
698            self,
699            expression: exp.TableSample,
700            sep: str = " AS ",
701            tablesample_keyword: t.Optional[str] = None,
702        ) -> str:
703            if not isinstance(expression.parent, exp.Select):
704                # This sample clause only applies to a single source, not the entire resulting relation
705                tablesample_keyword = "TABLESAMPLE"
706
707            if expression.args.get("size"):
708                method = expression.args.get("method")
709                if method and method.name.upper() != "RESERVOIR":
710                    self.unsupported(
711                        f"Sampling method {method} is not supported with a discrete sample count, "
712                        "defaulting to reservoir sampling"
713                    )
714                    expression.set("method", exp.var("RESERVOIR"))
715
716            return super().tablesample_sql(
717                expression, sep=sep, tablesample_keyword=tablesample_keyword
718            )
719
720        def interval_sql(self, expression: exp.Interval) -> str:
721            multiplier: t.Optional[int] = None
722            unit = expression.text("unit").lower()
723
724            if unit.startswith("week"):
725                multiplier = 7
726            if unit.startswith("quarter"):
727                multiplier = 90
728
729            if multiplier:
730                return f"({multiplier} * {super().interval_sql(exp.Interval(this=expression.this, unit=exp.var('DAY')))})"
731
732            return super().interval_sql(expression)
733
734        def columndef_sql(self, expression: exp.ColumnDef, sep: str = " ") -> str:
735            if isinstance(expression.parent, exp.UserDefinedFunction):
736                return self.sql(expression, "this")
737            return super().columndef_sql(expression, sep)
738
739        def join_sql(self, expression: exp.Join) -> str:
740            if (
741                expression.side == "LEFT"
742                and not expression.args.get("on")
743                and isinstance(expression.this, exp.Unnest)
744            ):
745                # Some dialects support `LEFT JOIN UNNEST(...)` without an explicit ON clause
746                # DuckDB doesn't, but we can just add a dummy ON clause that is always true
747                return super().join_sql(expression.on(exp.true()))
748
749            return super().join_sql(expression)
750
751        def generateseries_sql(self, expression: exp.GenerateSeries) -> str:
752            # GENERATE_SERIES(a, b) -> [a, b], RANGE(a, b) -> [a, b)
753            if expression.args.get("is_end_exclusive"):
754                return rename_func("RANGE")(self, expression)
755
756            return self.function_fallback_sql(expression)
757
758        def bracket_sql(self, expression: exp.Bracket) -> str:
759            this = expression.this
760            if isinstance(this, exp.Array):
761                this.replace(exp.paren(this))
762
763            bracket = super().bracket_sql(expression)
764
765            if not expression.args.get("returns_list_for_maps"):
766                if not this.type:
767                    from sqlglot.optimizer.annotate_types import annotate_types
768
769                    this = annotate_types(this)
770
771                if this.is_type(exp.DataType.Type.MAP):
772                    bracket = f"({bracket})[1]"
773
774            return bracket
775
776        def withingroup_sql(self, expression: exp.WithinGroup) -> str:
777            expression_sql = self.sql(expression, "expression")
778
779            func = expression.this
780            if isinstance(func, exp.PERCENTILES):
781                # Make the order key the first arg and slide the fraction to the right
782                # https://duckdb.org/docs/sql/aggregates#ordered-set-aggregate-functions
783                order_col = expression.find(exp.Ordered)
784                if order_col:
785                    func.set("expression", func.this)
786                    func.set("this", order_col.this)
787
788            this = self.sql(expression, "this").rstrip(")")
789
790            return f"{this}{expression_sql})"
791
792        def length_sql(self, expression: exp.Length) -> str:
793            arg = expression.this
794
795            # Dialects like BQ and Snowflake also accept binary values as args, so
796            # DDB will attempt to infer the type or resort to case/when resolution
797            if not expression.args.get("binary") or arg.is_string:
798                return self.func("LENGTH", arg)
799
800            if not arg.type:
801                from sqlglot.optimizer.annotate_types import annotate_types
802
803                arg = annotate_types(arg)
804
805            if arg.is_type(*exp.DataType.TEXT_TYPES):
806                return self.func("LENGTH", arg)
807
808            # We need these casts to make duckdb's static type checker happy
809            blob = exp.cast(arg, exp.DataType.Type.VARBINARY)
810            varchar = exp.cast(arg, exp.DataType.Type.VARCHAR)
811
812            case = (
813                exp.case(self.func("TYPEOF", arg))
814                .when(
815                    "'VARCHAR'", exp.Anonymous(this="LENGTH", expressions=[varchar])
816                )  # anonymous to break length_sql recursion
817                .when("'BLOB'", self.func("OCTET_LENGTH", blob))
818            )
819
820            return self.sql(case)
821
822        def objectinsert_sql(self, expression: exp.ObjectInsert) -> str:
823            this = expression.this
824            key = expression.args.get("key")
825            key_sql = key.name if isinstance(key, exp.Expression) else ""
826            value_sql = self.sql(expression, "value")
827
828            kv_sql = f"{key_sql} := {value_sql}"
829
830            # If the input struct is empty e.g. transpiling OBJECT_INSERT(OBJECT_CONSTRUCT(), key, value) from Snowflake
831            # then we can generate STRUCT_PACK which will build it since STRUCT_INSERT({}, key := value) is not valid DuckDB
832            if isinstance(this, exp.Struct) and not this.expressions:
833                return self.func("STRUCT_PACK", kv_sql)
834
835            return self.func("STRUCT_INSERT", this, kv_sql)
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

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]:
213    def to_json_path(self, path: t.Optional[exp.Expression]) -> t.Optional[exp.Expression]:
214        if isinstance(path, exp.Literal):
215            # DuckDB also supports the JSON pointer syntax, where every path starts with a `/`.
216            # Additionally, it allows accessing the back of lists using the `[#-i]` syntax.
217            # This check ensures we'll avoid trying to parse these as JSON paths, which can
218            # either result in a noisy warning or in an invalid representation of the path.
219            path_text = path.name
220            if path_text.startswith("/") or "[#" in path_text:
221                return path
222
223        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 = {}
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):
225    class Tokenizer(tokens.Tokenizer):
226        HEREDOC_STRINGS = ["$"]
227
228        HEREDOC_TAG_IS_IDENTIFIER = True
229        HEREDOC_STRING_ALTERNATIVE = TokenType.PARAMETER
230
231        KEYWORDS = {
232            **tokens.Tokenizer.KEYWORDS,
233            "//": TokenType.DIV,
234            "ATTACH": TokenType.COMMAND,
235            "BINARY": TokenType.VARBINARY,
236            "BITSTRING": TokenType.BIT,
237            "BPCHAR": TokenType.TEXT,
238            "CHAR": TokenType.TEXT,
239            "CHARACTER VARYING": TokenType.TEXT,
240            "EXCLUDE": TokenType.EXCEPT,
241            "LOGICAL": TokenType.BOOLEAN,
242            "ONLY": TokenType.ONLY,
243            "PIVOT_WIDER": TokenType.PIVOT,
244            "POSITIONAL": TokenType.POSITIONAL,
245            "SIGNED": TokenType.INT,
246            "STRING": TokenType.TEXT,
247            "SUMMARIZE": TokenType.SUMMARIZE,
248            "TIMESTAMP_S": TokenType.TIMESTAMP_S,
249            "TIMESTAMP_MS": TokenType.TIMESTAMP_MS,
250            "TIMESTAMP_NS": TokenType.TIMESTAMP_NS,
251            "TIMESTAMP_US": TokenType.TIMESTAMP,
252            "UBIGINT": TokenType.UBIGINT,
253            "UINTEGER": TokenType.UINT,
254            "USMALLINT": TokenType.USMALLINT,
255            "UTINYINT": TokenType.UTINYINT,
256            "VARCHAR": TokenType.TEXT,
257        }
258        KEYWORDS.pop("/*+")
259
260        SINGLE_TOKENS = {
261            **tokens.Tokenizer.SINGLE_TOKENS,
262            "$": TokenType.PARAMETER,
263        }
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'>, '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'>, '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'>, 'INT2': <TokenType.SMALLINT: 'SMALLINT'>, 'INTEGER': <TokenType.INT: 'INT'>, 'INT': <TokenType.INT: 'INT'>, 'INT4': <TokenType.INT: 'INT'>, 'INT32': <TokenType.INT: 'INT'>, 'INT64': <TokenType.BIGINT: 'BIGINT'>, 'LONG': <TokenType.BIGINT: 'BIGINT'>, 'BIGINT': <TokenType.BIGINT: 'BIGINT'>, 'INT8': <TokenType.TINYINT: 'TINYINT'>, 'UINT': <TokenType.UINT: 'UINT'>, 'DEC': <TokenType.DECIMAL: 'DECIMAL'>, 'DECIMAL': <TokenType.DECIMAL: 'DECIMAL'>, 'BIGDECIMAL': <TokenType.BIGDECIMAL: 'BIGDECIMAL'>, 'BIGNUMERIC': <TokenType.BIGDECIMAL: 'BIGDECIMAL'>, 'LIST': <TokenType.LIST: 'LIST'>, 'MAP': <TokenType.MAP: 'MAP'>, 'NULLABLE': <TokenType.NULLABLE: 'NULLABLE'>, 'NUMBER': <TokenType.DECIMAL: 'DECIMAL'>, 'NUMERIC': <TokenType.DECIMAL: 'DECIMAL'>, 'FIXED': <TokenType.DECIMAL: 'DECIMAL'>, 'REAL': <TokenType.FLOAT: 'FLOAT'>, 'FLOAT': <TokenType.FLOAT: 'FLOAT'>, 'FLOAT4': <TokenType.FLOAT: 'FLOAT'>, 'FLOAT8': <TokenType.DOUBLE: 'DOUBLE'>, 'DOUBLE': <TokenType.DOUBLE: 'DOUBLE'>, 'DOUBLE PRECISION': <TokenType.DOUBLE: 'DOUBLE'>, 'JSON': <TokenType.JSON: 'JSON'>, 'JSONB': <TokenType.JSONB: 'JSONB'>, 'CHAR': <TokenType.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.COMMAND: 'COMMAND'>, 'OPTIMIZE': <TokenType.COMMAND: 'COMMAND'>, 'PREPARE': <TokenType.COMMAND: 'COMMAND'>, 'VACUUM': <TokenType.COMMAND: 'COMMAND'>, 'USER-DEFINED': <TokenType.USERDEFINED: 'USERDEFINED'>, 'FOR VERSION': <TokenType.VERSION_SNAPSHOT: 'VERSION_SNAPSHOT'>, 'FOR TIMESTAMP': <TokenType.TIMESTAMP_SNAPSHOT: 'TIMESTAMP_SNAPSHOT'>, '//': <TokenType.DIV: 'DIV'>, '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):
265    class Parser(parser.Parser):
266        BITWISE = {
267            **parser.Parser.BITWISE,
268            TokenType.TILDA: exp.RegexpLike,
269        }
270
271        FUNCTIONS_WITH_ALIASED_ARGS = {*parser.Parser.FUNCTIONS_WITH_ALIASED_ARGS, "STRUCT_PACK"}
272
273        FUNCTIONS = {
274            **parser.Parser.FUNCTIONS,
275            "ARRAY_REVERSE_SORT": _build_sort_array_desc,
276            "ARRAY_SORT": exp.SortArray.from_arg_list,
277            "DATEDIFF": _build_date_diff,
278            "DATE_DIFF": _build_date_diff,
279            "DATE_TRUNC": date_trunc_to_time,
280            "DATETRUNC": date_trunc_to_time,
281            "DECODE": lambda args: exp.Decode(
282                this=seq_get(args, 0), charset=exp.Literal.string("utf-8")
283            ),
284            "ENCODE": lambda args: exp.Encode(
285                this=seq_get(args, 0), charset=exp.Literal.string("utf-8")
286            ),
287            "EPOCH": exp.TimeToUnix.from_arg_list,
288            "EPOCH_MS": lambda args: exp.UnixToTime(
289                this=seq_get(args, 0), scale=exp.UnixToTime.MILLIS
290            ),
291            "JSON": exp.ParseJSON.from_arg_list,
292            "JSON_EXTRACT_PATH": parser.build_extract_json_with_path(exp.JSONExtract),
293            "JSON_EXTRACT_STRING": parser.build_extract_json_with_path(exp.JSONExtractScalar),
294            "LIST_HAS": exp.ArrayContains.from_arg_list,
295            "LIST_REVERSE_SORT": _build_sort_array_desc,
296            "LIST_SORT": exp.SortArray.from_arg_list,
297            "LIST_VALUE": exp.Array.from_arg_list,
298            "MAKE_TIME": exp.TimeFromParts.from_arg_list,
299            "MAKE_TIMESTAMP": _build_make_timestamp,
300            "MEDIAN": lambda args: exp.PercentileCont(
301                this=seq_get(args, 0), expression=exp.Literal.number(0.5)
302            ),
303            "QUANTILE_CONT": exp.PercentileCont.from_arg_list,
304            "QUANTILE_DISC": exp.PercentileDisc.from_arg_list,
305            "REGEXP_EXTRACT": lambda args: exp.RegexpExtract(
306                this=seq_get(args, 0), expression=seq_get(args, 1), group=seq_get(args, 2)
307            ),
308            "REGEXP_MATCHES": exp.RegexpLike.from_arg_list,
309            "REGEXP_REPLACE": lambda args: exp.RegexpReplace(
310                this=seq_get(args, 0),
311                expression=seq_get(args, 1),
312                replacement=seq_get(args, 2),
313                modifiers=seq_get(args, 3),
314            ),
315            "STRFTIME": build_formatted_time(exp.TimeToStr, "duckdb"),
316            "STRING_SPLIT": exp.Split.from_arg_list,
317            "STRING_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
318            "STRING_TO_ARRAY": exp.Split.from_arg_list,
319            "STRPTIME": build_formatted_time(exp.StrToTime, "duckdb"),
320            "STRUCT_PACK": exp.Struct.from_arg_list,
321            "STR_SPLIT": exp.Split.from_arg_list,
322            "STR_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
323            "TO_TIMESTAMP": exp.UnixToTime.from_arg_list,
324            "UNNEST": exp.Explode.from_arg_list,
325            "XOR": binary_from_function(exp.BitwiseXor),
326            "GENERATE_SERIES": _build_generate_series(),
327            "RANGE": _build_generate_series(end_exclusive=True),
328        }
329
330        FUNCTIONS.pop("DATE_SUB")
331
332        FUNCTION_PARSERS = parser.Parser.FUNCTION_PARSERS.copy()
333        FUNCTION_PARSERS.pop("DECODE")
334
335        NO_PAREN_FUNCTION_PARSERS = {
336            **parser.Parser.NO_PAREN_FUNCTION_PARSERS,
337            "MAP": lambda self: self._parse_map(),
338        }
339
340        TABLE_ALIAS_TOKENS = parser.Parser.TABLE_ALIAS_TOKENS - {
341            TokenType.SEMI,
342            TokenType.ANTI,
343        }
344
345        PLACEHOLDER_PARSERS = {
346            **parser.Parser.PLACEHOLDER_PARSERS,
347            TokenType.PARAMETER: lambda self: (
348                self.expression(exp.Placeholder, this=self._prev.text)
349                if self._match(TokenType.NUMBER) or self._match_set(self.ID_VAR_TOKENS)
350                else None
351            ),
352        }
353
354        TYPE_CONVERTERS = {
355            # https://duckdb.org/docs/sql/data_types/numeric
356            exp.DataType.Type.DECIMAL: build_default_decimal_type(precision=18, scale=3),
357            # https://duckdb.org/docs/sql/data_types/text
358            exp.DataType.Type.TEXT: lambda dtype: exp.DataType.build("TEXT"),
359        }
360
361        def _parse_table_sample(self, as_modifier: bool = False) -> t.Optional[exp.TableSample]:
362            # https://duckdb.org/docs/sql/samples.html
363            sample = super()._parse_table_sample(as_modifier=as_modifier)
364            if sample and not sample.args.get("method"):
365                if sample.args.get("size"):
366                    sample.set("method", exp.var("RESERVOIR"))
367                else:
368                    sample.set("method", exp.var("SYSTEM"))
369
370            return sample
371
372        def _parse_bracket(
373            self, this: t.Optional[exp.Expression] = None
374        ) -> t.Optional[exp.Expression]:
375            bracket = super()._parse_bracket(this)
376            if isinstance(bracket, exp.Bracket):
377                bracket.set("returns_list_for_maps", True)
378
379            return bracket
380
381        def _parse_map(self) -> exp.ToMap | exp.Map:
382            if self._match(TokenType.L_BRACE, advance=False):
383                return self.expression(exp.ToMap, this=self._parse_bracket())
384
385            args = self._parse_wrapped_csv(self._parse_assignment)
386            return self.expression(exp.Map, keys=seq_get(args, 0), values=seq_get(args, 1))
387
388        def _parse_struct_types(self, type_required: bool = False) -> t.Optional[exp.Expression]:
389            return self._parse_field_def()
390
391        def _pivot_column_names(self, aggregations: t.List[exp.Expression]) -> t.List[str]:
392            if len(aggregations) == 1:
393                return super()._pivot_column_names(aggregations)
394            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.CARET: 'CARET'>: <class 'sqlglot.expressions.BitwiseXor'>, <TokenType.PIPE: 'PIPE'>: <class 'sqlglot.expressions.BitwiseOr'>, <TokenType.TILDA: 'TILDA'>: <class 'sqlglot.expressions.RegexpLike'>}
FUNCTIONS_WITH_ALIASED_ARGS = {'STRUCT_PACK', 'STRUCT'}
FUNCTIONS = {'ABS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Abs'>>, 'ADD_MONTHS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.AddMonths'>>, 'ANONYMOUS_AGG_FUNC': <bound method Func.from_arg_list of <class 'sqlglot.expressions.AnonymousAggFunc'>>, 'ANY_VALUE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.AnyValue'>>, 'APPROX_DISTINCT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ApproxDistinct'>>, 'APPROX_COUNT_DISTINCT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ApproxDistinct'>>, 'APPROX_QUANTILE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ApproxQuantile'>>, 'APPROX_TOP_K': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ApproxTopK'>>, 'ARG_MAX': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArgMax'>>, 'ARGMAX': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArgMax'>>, 'MAX_BY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArgMax'>>, 'ARG_MIN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArgMin'>>, 'ARGMIN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArgMin'>>, 'MIN_BY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArgMin'>>, 'ARRAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Array'>>, 'ARRAY_AGG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayAgg'>>, 'ARRAY_ALL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayAll'>>, 'ARRAY_ANY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayAny'>>, 'ARRAY_CONCAT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayConcat'>>, 'ARRAY_CAT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayConcat'>>, 'ARRAY_CONSTRUCT_COMPACT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayConstructCompact'>>, 'ARRAY_CONTAINS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayContains'>>, 'ARRAY_HAS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayContains'>>, 'ARRAY_CONTAINS_ALL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayContainsAll'>>, 'ARRAY_HAS_ALL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayContainsAll'>>, 'FILTER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayFilter'>>, 'ARRAY_FILTER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayFilter'>>, 'ARRAY_OVERLAPS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayOverlaps'>>, 'ARRAY_SIZE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArraySize'>>, 'ARRAY_LENGTH': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArraySize'>>, 'ARRAY_SORT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.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': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Chr'>>, 'CHAR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Chr'>>, 'COALESCE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Coalesce'>>, 'IFNULL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Coalesce'>>, 'NVL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Coalesce'>>, 'COLLATE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Collate'>>, 'COMBINED_AGG_FUNC': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CombinedAggFunc'>>, 'COMBINED_PARAMETERIZED_AGG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CombinedParameterizedAgg'>>, 'CONCAT': <function Parser.<lambda>>, 'CONCAT_WS': <function Parser.<lambda>>, 'CONNECT_BY_ROOT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ConnectByRoot'>>, 'CONVERT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Convert'>>, 'CORR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Corr'>>, 'COUNT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Count'>>, 'COUNT_IF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CountIf'>>, 'COUNTIF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CountIf'>>, 'COVAR_POP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CovarPop'>>, 'COVAR_SAMP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CovarSamp'>>, 'CURRENT_DATE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CurrentDate'>>, 'CURRENT_DATETIME': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CurrentDatetime'>>, 'CURRENT_TIME': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CurrentTime'>>, 'CURRENT_TIMESTAMP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CurrentTimestamp'>>, 'CURRENT_USER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.CurrentUser'>>, 'DATE': <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'>>, '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>>, 'EXP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Exp'>>, 'EXPLODE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Explode'>>, 'EXPLODE_OUTER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ExplodeOuter'>>, 'EXTRACT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Extract'>>, 'FIRST': <bound method Func.from_arg_list of <class 'sqlglot.expressions.First'>>, 'FIRST_VALUE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.FirstValue'>>, 'FLATTEN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Flatten'>>, 'FLOOR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Floor'>>, 'FROM_BASE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.FromBase'>>, 'FROM_BASE64': <bound method Func.from_arg_list of <class 'sqlglot.expressions.FromBase64'>>, 'GAP_FILL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.GapFill'>>, 'GENERATE_DATE_ARRAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.GenerateDateArray'>>, 'GENERATE_SERIES': <function _build_generate_series.<locals>._builder>, 'GREATEST': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Greatest'>>, 'GROUP_CONCAT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.GroupConcat'>>, 'HEX': <function build_hex>, 'HLL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Hll'>>, 'IF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.If'>>, 'IIF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.If'>>, 'INITCAP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Initcap'>>, 'IS_INF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.IsInf'>>, 'ISINF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.IsInf'>>, 'IS_NAN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.IsNan'>>, 'ISNAN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.IsNan'>>, 'J_S_O_N_ARRAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONArray'>>, 'J_S_O_N_ARRAY_AGG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONArrayAgg'>>, 'JSON_ARRAY_CONTAINS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONArrayContains'>>, 'JSONB_CONTAINS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONBContains'>>, 'JSONB_EXTRACT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONBExtract'>>, 'JSONB_EXTRACT_SCALAR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONBExtractScalar'>>, 'JSON_EXTRACT': <function build_extract_json_with_path.<locals>._builder>, 'JSON_EXTRACT_SCALAR': <function 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'>>, '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'>>, 'MIN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Min'>>, 'MONTH': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Month'>>, 'MONTHS_BETWEEN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.MonthsBetween'>>, 'NEXT_VALUE_FOR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.NextValueFor'>>, 'NTH_VALUE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.NthValue'>>, 'NULLIF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Nullif'>>, 'NUMBER_TO_STR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.NumberToStr'>>, 'NVL2': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Nvl2'>>, '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'>>, '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 DuckDB.Parser.<lambda>>, 'REGEXP_I_LIKE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.RegexpILike'>>, 'REGEXP_LIKE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.RegexpLike'>>, 'REGEXP_REPLACE': <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'>>, '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_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'>>, '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_MAP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ToMap'>>, 'TO_NUMBER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ToNumber'>>, 'TRANSFORM': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Transform'>>, 'TRIM': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Trim'>>, 'TRY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Try'>>, 'TRY_CAST': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TryCast'>>, 'TS_OR_DI_TO_DI': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDiToDi'>>, 'TS_OR_DS_ADD': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDsAdd'>>, 'TS_OR_DS_DIFF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDsDiff'>>, 'TS_OR_DS_TO_DATE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDsToDate'>>, 'TS_OR_DS_TO_DATE_STR': <function Parser.<lambda>>, 'TS_OR_DS_TO_TIME': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDsToTime'>>, 'TS_OR_DS_TO_TIMESTAMP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDsToTimestamp'>>, 'UNHEX': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Unhex'>>, 'UNIX_DATE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.UnixDate'>>, 'UNIX_TO_STR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.UnixToStr'>>, 'UNIX_TO_TIME': <bound method Func.from_arg_list of <class 'sqlglot.expressions.UnixToTime'>>, 'UNIX_TO_TIME_STR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.UnixToTimeStr'>>, 'UNNEST': <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'>>, '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'>>, 'GLOB': <function Parser.<lambda>>, 'JSON_EXTRACT_PATH_TEXT': <function build_extract_json_with_path.<locals>._builder>, 'LIKE': <function build_like>, 'LOG2': <function Parser.<lambda>>, 'LOG10': <function Parser.<lambda>>, 'LPAD': <function Parser.<lambda>>, 'LEFTPAD': <function Parser.<lambda>>, 'MOD': <function build_mod>, 'RPAD': <function Parser.<lambda>>, 'RIGHTPAD': <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': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Array'>>, 'MAKE_TIME': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TimeFromParts'>>, 'MAKE_TIMESTAMP': <function _build_make_timestamp>, 'MEDIAN': <function DuckDB.Parser.<lambda>>, '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>}
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>>, 'OPENJSON': <function Parser.<lambda>>, 'POSITION': <function Parser.<lambda>>, 'PREDICT': <function Parser.<lambda>>, 'SAFE_CAST': <function Parser.<lambda>>, 'STRING_AGG': <function Parser.<lambda>>, 'SUBSTRING': <function Parser.<lambda>>, '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.ESCAPE: 'ESCAPE'>, <TokenType.REPLACE: 'REPLACE'>, <TokenType.DESC: 'DESC'>, <TokenType.TOP: 'TOP'>, <TokenType.USE: 'USE'>, <TokenType.DATERANGE: 'DATERANGE'>, <TokenType.TSMULTIRANGE: 'TSMULTIRANGE'>, <TokenType.KEEP: 'KEEP'>, <TokenType.TINYBLOB: 'TINYBLOB'>, <TokenType.TIMESTAMP_NS: 'TIMESTAMP_NS'>, <TokenType.UINT256: 'UINT256'>, <TokenType.LOAD: 'LOAD'>, <TokenType.EXISTS: 'EXISTS'>, <TokenType.SOME: 'SOME'>, <TokenType.BIGSERIAL: 'BIGSERIAL'>, <TokenType.OBJECT_IDENTIFIER: 'OBJECT_IDENTIFIER'>, <TokenType.FILTER: 'FILTER'>, <TokenType.DATABASE: 'DATABASE'>, <TokenType.BIGINT: 'BIGINT'>, <TokenType.SMALLINT: 'SMALLINT'>, <TokenType.XML: 'XML'>, <TokenType.DIV: 'DIV'>, <TokenType.UUID: 'UUID'>, <TokenType.MONEY: 'MONEY'>, <TokenType.ENUM: 'ENUM'>, <TokenType.COLLATE: 'COLLATE'>, <TokenType.KILL: 'KILL'>, <TokenType.TIMESTAMP_MS: 'TIMESTAMP_MS'>, <TokenType.VARIANT: 'VARIANT'>, <TokenType.ARRAY: 'ARRAY'>, <TokenType.DELETE: 'DELETE'>, <TokenType.DATE: 'DATE'>, <TokenType.NULLABLE: 'NULLABLE'>, <TokenType.STRUCT: 'STRUCT'>, <TokenType.INT4MULTIRANGE: 'INT4MULTIRANGE'>, <TokenType.FALSE: 'FALSE'>, <TokenType.ORDINALITY: 'ORDINALITY'>, <TokenType.BOOLEAN: 'BOOLEAN'>, <TokenType.IS: 'IS'>, <TokenType.FIRST: 'FIRST'>, <TokenType.STORAGE_INTEGRATION: 'STORAGE_INTEGRATION'>, <TokenType.BEGIN: 'BEGIN'>, <TokenType.HSTORE: 'HSTORE'>, <TokenType.CHAR: 'CHAR'>, <TokenType.ROWS: 'ROWS'>, <TokenType.SUPER: 'SUPER'>, <TokenType.INT4RANGE: 'INT4RANGE'>, <TokenType.TSTZMULTIRANGE: 'TSTZMULTIRANGE'>, <TokenType.INT: 'INT'>, <TokenType.DATE32: 'DATE32'>, <TokenType.SETTINGS: 'SETTINGS'>, <TokenType.MODEL: 'MODEL'>, <TokenType.IDENTIFIER: 'IDENTIFIER'>, <TokenType.MEDIUMBLOB: 'MEDIUMBLOB'>, <TokenType.UINT128: 'UINT128'>, <TokenType.SCHEMA: 'SCHEMA'>, <TokenType.IPV6: 'IPV6'>, <TokenType.COPY: 'COPY'>, <TokenType.UNPIVOT: 'UNPIVOT'>, <TokenType.BPCHAR: 'BPCHAR'>, <TokenType.GEOMETRY: 'GEOMETRY'>, <TokenType.DATETIME: 'DATETIME'>, <TokenType.TSTZRANGE: 'TSTZRANGE'>, <TokenType.CASE: 'CASE'>, <TokenType.VIEW: 'VIEW'>, <TokenType.IPPREFIX: 'IPPREFIX'>, <TokenType.DATETIME64: 'DATETIME64'>, <TokenType.FINAL: 'FINAL'>, <TokenType.NUMMULTIRANGE: 'NUMMULTIRANGE'>, <TokenType.IPADDRESS: 'IPADDRESS'>, <TokenType.TINYTEXT: 'TINYTEXT'>, <TokenType.CURRENT_TIMESTAMP: 'CURRENT_TIMESTAMP'>, <TokenType.SHOW: 'SHOW'>, <TokenType.NEXT: 'NEXT'>, <TokenType.RECURSIVE: 'RECURSIVE'>, <TokenType.PERCENT: 'PERCENT'>, <TokenType.COLUMN: 'COLUMN'>, <TokenType.REFERENCES: 'REFERENCES'>, <TokenType.OVERWRITE: 'OVERWRITE'>, <TokenType.OPERATOR: 'OPERATOR'>, <TokenType.DEFAULT: 'DEFAULT'>, <TokenType.JSON: 'JSON'>, <TokenType.TSRANGE: 'TSRANGE'>, <TokenType.COMMAND: 'COMMAND'>, <TokenType.TEMPORARY: 'TEMPORARY'>, <TokenType.CURRENT_DATE: 'CURRENT_DATE'>, <TokenType.TEXT: 'TEXT'>, <TokenType.SMALLSERIAL: 'SMALLSERIAL'>, <TokenType.HLLSKETCH: 'HLLSKETCH'>, <TokenType.SET: 'SET'>, <TokenType.UNKNOWN: 'UNKNOWN'>, <TokenType.VAR: 'VAR'>, <TokenType.ASC: 'ASC'>, <TokenType.PARTITION: 'PARTITION'>, <TokenType.SMALLMONEY: 'SMALLMONEY'>, <TokenType.DATEMULTIRANGE: 'DATEMULTIRANGE'>, <TokenType.TIME: 'TIME'>, <TokenType.LONGTEXT: 'LONGTEXT'>, <TokenType.ALL: 'ALL'>, <TokenType.MEDIUMTEXT: 'MEDIUMTEXT'>, <TokenType.CUBE: 'CUBE'>, <TokenType.ENUM8: 'ENUM8'>, <TokenType.UMEDIUMINT: 'UMEDIUMINT'>, <TokenType.ENUM16: 'ENUM16'>, <TokenType.ROWVERSION: 'ROWVERSION'>, <TokenType.BINARY: 'BINARY'>, <TokenType.NUMRANGE: 'NUMRANGE'>, <TokenType.TABLE: 'TABLE'>, <TokenType.UTINYINT: 'UTINYINT'>, <TokenType.GEOGRAPHY: 'GEOGRAPHY'>, <TokenType.PROCEDURE: 'PROCEDURE'>, <TokenType.UNNEST: 'UNNEST'>, <TokenType.UNIQUEIDENTIFIER: 'UNIQUEIDENTIFIER'>, <TokenType.UBIGINT: 'UBIGINT'>, <TokenType.DICTIONARY: 'DICTIONARY'>, <TokenType.INT256: 'INT256'>, <TokenType.COMMENT: 'COMMENT'>, <TokenType.MERGE: 'MERGE'>, <TokenType.SERIAL: 'SERIAL'>, <TokenType.IPV4: 'IPV4'>, <TokenType.BIGDECIMAL: 'BIGDECIMAL'>, <TokenType.TINYINT: 'TINYINT'>, <TokenType.CURRENT_DATETIME: 'CURRENT_DATETIME'>, <TokenType.UNIQUE: 'UNIQUE'>, <TokenType.LONGBLOB: 'LONGBLOB'>, <TokenType.FORMAT: 'FORMAT'>, <TokenType.COMMIT: 'COMMIT'>, <TokenType.BIT: 'BIT'>, <TokenType.OBJECT: 'OBJECT'>, <TokenType.STRAIGHT_JOIN: 'STRAIGHT_JOIN'>, <TokenType.TIMESTAMPNTZ: 'TIMESTAMPNTZ'>, <TokenType.VARBINARY: 'VARBINARY'>, <TokenType.LIST: 'LIST'>, <TokenType.FLOAT: 'FLOAT'>, <TokenType.DECIMAL: 'DECIMAL'>, <TokenType.TRUNCATE: 'TRUNCATE'>, <TokenType.RANGE: 'RANGE'>, <TokenType.PIVOT: 'PIVOT'>, <TokenType.TIMESTAMPTZ: 'TIMESTAMPTZ'>, <TokenType.YEAR: 'YEAR'>, <TokenType.NESTED: 'NESTED'>, <TokenType.VOLATILE: 'VOLATILE'>, <TokenType.INT8RANGE: 'INT8RANGE'>, <TokenType.INT128: 'INT128'>, <TokenType.TRUE: 'TRUE'>, <TokenType.MAP: 'MAP'>, <TokenType.INT8MULTIRANGE: 'INT8MULTIRANGE'>, <TokenType.USERDEFINED: 'USERDEFINED'>, <TokenType.LOWCARDINALITY: 'LOWCARDINALITY'>, <TokenType.VARCHAR: 'VARCHAR'>, <TokenType.FIXEDSTRING: 'FIXEDSTRING'>, <TokenType.INDEX: 'INDEX'>, <TokenType.AUTO_INCREMENT: 'AUTO_INCREMENT'>, <TokenType.CURRENT_USER: 'CURRENT_USER'>, <TokenType.JSONB: 'JSONB'>, <TokenType.USMALLINT: 'USMALLINT'>, <TokenType.CONSTRAINT: 'CONSTRAINT'>, <TokenType.INTERVAL: 'INTERVAL'>, <TokenType.REFRESH: 'REFRESH'>, <TokenType.STREAMLIT: 'STREAMLIT'>, <TokenType.END: 'END'>, <TokenType.FOREIGN_KEY: 'FOREIGN_KEY'>, <TokenType.DESCRIBE: 'DESCRIBE'>, <TokenType.UINT: 'UINT'>, <TokenType.TDIGEST: 'TDIGEST'>, <TokenType.NAME: 'NAME'>, <TokenType.OVERLAPS: 'OVERLAPS'>, <TokenType.TIMESTAMPLTZ: 'TIMESTAMPLTZ'>, <TokenType.TIMETZ: 'TIMETZ'>, <TokenType.PRAGMA: 'PRAGMA'>, <TokenType.ROW: 'ROW'>, <TokenType.ROLLUP: 'ROLLUP'>, <TokenType.INET: 'INET'>, <TokenType.FUNCTION: 'FUNCTION'>, <TokenType.ANY: 'ANY'>, <TokenType.NCHAR: 'NCHAR'>, <TokenType.NVARCHAR: 'NVARCHAR'>, <TokenType.VECTOR: 'VECTOR'>, <TokenType.UDECIMAL: 'UDECIMAL'>, <TokenType.EXECUTE: 'EXECUTE'>, <TokenType.TIMESTAMP_S: 'TIMESTAMP_S'>, <TokenType.TIMESTAMP: 'TIMESTAMP'>, <TokenType.NULL: 'NULL'>, <TokenType.PSEUDO_TYPE: 'PSEUDO_TYPE'>, <TokenType.UPDATE: 'UPDATE'>, <TokenType.TAG: 'TAG'>, <TokenType.IMAGE: 'IMAGE'>, <TokenType.CACHE: 'CACHE'>, <TokenType.SEQUENCE: 'SEQUENCE'>, <TokenType.MEDIUMINT: 'MEDIUMINT'>, <TokenType.SIMPLEAGGREGATEFUNCTION: 'SIMPLEAGGREGATEFUNCTION'>, <TokenType.ISNULL: 'ISNULL'>, <TokenType.DOUBLE: 'DOUBLE'>, <TokenType.AGGREGATEFUNCTION: 'AGGREGATEFUNCTION'>, <TokenType.WAREHOUSE: 'WAREHOUSE'>, <TokenType.CURRENT_TIME: 'CURRENT_TIME'>}
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.ESCAPE: 'ESCAPE'>, <TokenType.REPLACE: 'REPLACE'>, <TokenType.DESC: 'DESC'>, <TokenType.TOP: 'TOP'>, <TokenType.USE: 'USE'>, <TokenType.DATERANGE: 'DATERANGE'>, <TokenType.TSMULTIRANGE: 'TSMULTIRANGE'>, <TokenType.KEEP: 'KEEP'>, <TokenType.TINYBLOB: 'TINYBLOB'>, <TokenType.TIMESTAMP_NS: 'TIMESTAMP_NS'>, <TokenType.UINT256: 'UINT256'>, <TokenType.ANTI: 'ANTI'>, <TokenType.LOAD: 'LOAD'>, <TokenType.EXISTS: 'EXISTS'>, <TokenType.SOME: 'SOME'>, <TokenType.BIGSERIAL: 'BIGSERIAL'>, <TokenType.OBJECT_IDENTIFIER: 'OBJECT_IDENTIFIER'>, <TokenType.FILTER: 'FILTER'>, <TokenType.DATABASE: 'DATABASE'>, <TokenType.BIGINT: 'BIGINT'>, <TokenType.SMALLINT: 'SMALLINT'>, <TokenType.XML: 'XML'>, <TokenType.DIV: 'DIV'>, <TokenType.UUID: 'UUID'>, <TokenType.RIGHT: 'RIGHT'>, <TokenType.MONEY: 'MONEY'>, <TokenType.ENUM: 'ENUM'>, <TokenType.COLLATE: 'COLLATE'>, <TokenType.KILL: 'KILL'>, <TokenType.TIMESTAMP_MS: 'TIMESTAMP_MS'>, <TokenType.VARIANT: 'VARIANT'>, <TokenType.ARRAY: 'ARRAY'>, <TokenType.DELETE: 'DELETE'>, <TokenType.DATE: 'DATE'>, <TokenType.NULLABLE: 'NULLABLE'>, <TokenType.STRUCT: 'STRUCT'>, <TokenType.INT4MULTIRANGE: 'INT4MULTIRANGE'>, <TokenType.FALSE: 'FALSE'>, <TokenType.ORDINALITY: 'ORDINALITY'>, <TokenType.ASOF: 'ASOF'>, <TokenType.IS: 'IS'>, <TokenType.BOOLEAN: 'BOOLEAN'>, <TokenType.FIRST: 'FIRST'>, <TokenType.STORAGE_INTEGRATION: 'STORAGE_INTEGRATION'>, <TokenType.BEGIN: 'BEGIN'>, <TokenType.HSTORE: 'HSTORE'>, <TokenType.CHAR: 'CHAR'>, <TokenType.ROWS: 'ROWS'>, <TokenType.SUPER: 'SUPER'>, <TokenType.INT4RANGE: 'INT4RANGE'>, <TokenType.TSTZMULTIRANGE: 'TSTZMULTIRANGE'>, <TokenType.INT: 'INT'>, <TokenType.DATE32: 'DATE32'>, <TokenType.SETTINGS: 'SETTINGS'>, <TokenType.MODEL: 'MODEL'>, <TokenType.IDENTIFIER: 'IDENTIFIER'>, <TokenType.MEDIUMBLOB: 'MEDIUMBLOB'>, <TokenType.UINT128: 'UINT128'>, <TokenType.SCHEMA: 'SCHEMA'>, <TokenType.IPV6: 'IPV6'>, <TokenType.COPY: 'COPY'>, <TokenType.SEMI: 'SEMI'>, <TokenType.BPCHAR: 'BPCHAR'>, <TokenType.UNPIVOT: 'UNPIVOT'>, <TokenType.GEOMETRY: 'GEOMETRY'>, <TokenType.DATETIME: 'DATETIME'>, <TokenType.TSTZRANGE: 'TSTZRANGE'>, <TokenType.CASE: 'CASE'>, <TokenType.VIEW: 'VIEW'>, <TokenType.IPPREFIX: 'IPPREFIX'>, <TokenType.DATETIME64: 'DATETIME64'>, <TokenType.FINAL: 'FINAL'>, <TokenType.NUMMULTIRANGE: 'NUMMULTIRANGE'>, <TokenType.IPADDRESS: 'IPADDRESS'>, <TokenType.TINYTEXT: 'TINYTEXT'>, <TokenType.CURRENT_TIMESTAMP: 'CURRENT_TIMESTAMP'>, <TokenType.SHOW: 'SHOW'>, <TokenType.NEXT: 'NEXT'>, <TokenType.RECURSIVE: 'RECURSIVE'>, <TokenType.PERCENT: 'PERCENT'>, <TokenType.COLUMN: 'COLUMN'>, <TokenType.REFERENCES: 'REFERENCES'>, <TokenType.OVERWRITE: 'OVERWRITE'>, <TokenType.OPERATOR: 'OPERATOR'>, <TokenType.DEFAULT: 'DEFAULT'>, <TokenType.JSON: 'JSON'>, <TokenType.TSRANGE: 'TSRANGE'>, <TokenType.COMMAND: 'COMMAND'>, <TokenType.TEMPORARY: 'TEMPORARY'>, <TokenType.CURRENT_DATE: 'CURRENT_DATE'>, <TokenType.TEXT: 'TEXT'>, <TokenType.SMALLSERIAL: 'SMALLSERIAL'>, <TokenType.HLLSKETCH: 'HLLSKETCH'>, <TokenType.SET: 'SET'>, <TokenType.UNKNOWN: 'UNKNOWN'>, <TokenType.VAR: 'VAR'>, <TokenType.ASC: 'ASC'>, <TokenType.PARTITION: 'PARTITION'>, <TokenType.SMALLMONEY: 'SMALLMONEY'>, <TokenType.DATEMULTIRANGE: 'DATEMULTIRANGE'>, <TokenType.TIME: 'TIME'>, <TokenType.LONGTEXT: 'LONGTEXT'>, <TokenType.ALL: 'ALL'>, <TokenType.MEDIUMTEXT: 'MEDIUMTEXT'>, <TokenType.CUBE: 'CUBE'>, <TokenType.ENUM8: 'ENUM8'>, <TokenType.UMEDIUMINT: 'UMEDIUMINT'>, <TokenType.ENUM16: 'ENUM16'>, <TokenType.ROWVERSION: 'ROWVERSION'>, <TokenType.BINARY: 'BINARY'>, <TokenType.NUMRANGE: 'NUMRANGE'>, <TokenType.TABLE: 'TABLE'>, <TokenType.UTINYINT: 'UTINYINT'>, <TokenType.GEOGRAPHY: 'GEOGRAPHY'>, <TokenType.PROCEDURE: 'PROCEDURE'>, <TokenType.UNNEST: 'UNNEST'>, <TokenType.UNIQUEIDENTIFIER: 'UNIQUEIDENTIFIER'>, <TokenType.UBIGINT: 'UBIGINT'>, <TokenType.DICTIONARY: 'DICTIONARY'>, <TokenType.INT256: 'INT256'>, <TokenType.COMMENT: 'COMMENT'>, <TokenType.MERGE: 'MERGE'>, <TokenType.WINDOW: 'WINDOW'>, <TokenType.SERIAL: 'SERIAL'>, <TokenType.IPV4: 'IPV4'>, <TokenType.BIGDECIMAL: 'BIGDECIMAL'>, <TokenType.TINYINT: 'TINYINT'>, <TokenType.CURRENT_DATETIME: 'CURRENT_DATETIME'>, <TokenType.UNIQUE: 'UNIQUE'>, <TokenType.LONGBLOB: 'LONGBLOB'>, <TokenType.FORMAT: 'FORMAT'>, <TokenType.COMMIT: 'COMMIT'>, <TokenType.NATURAL: 'NATURAL'>, <TokenType.BIT: 'BIT'>, <TokenType.OBJECT: 'OBJECT'>, <TokenType.STRAIGHT_JOIN: 'STRAIGHT_JOIN'>, <TokenType.TIMESTAMPNTZ: 'TIMESTAMPNTZ'>, <TokenType.VARBINARY: 'VARBINARY'>, <TokenType.LIST: 'LIST'>, <TokenType.FLOAT: 'FLOAT'>, <TokenType.DECIMAL: 'DECIMAL'>, <TokenType.TRUNCATE: 'TRUNCATE'>, <TokenType.RANGE: 'RANGE'>, <TokenType.PIVOT: 'PIVOT'>, <TokenType.TIMESTAMPTZ: 'TIMESTAMPTZ'>, <TokenType.YEAR: 'YEAR'>, <TokenType.NESTED: 'NESTED'>, <TokenType.VOLATILE: 'VOLATILE'>, <TokenType.INT8RANGE: 'INT8RANGE'>, <TokenType.INT128: 'INT128'>, <TokenType.TRUE: 'TRUE'>, <TokenType.MAP: 'MAP'>, <TokenType.INT8MULTIRANGE: 'INT8MULTIRANGE'>, <TokenType.USERDEFINED: 'USERDEFINED'>, <TokenType.LOWCARDINALITY: 'LOWCARDINALITY'>, <TokenType.VARCHAR: 'VARCHAR'>, <TokenType.FIXEDSTRING: 'FIXEDSTRING'>, <TokenType.INDEX: 'INDEX'>, <TokenType.LEFT: 'LEFT'>, <TokenType.AUTO_INCREMENT: 'AUTO_INCREMENT'>, <TokenType.CURRENT_USER: 'CURRENT_USER'>, <TokenType.JSONB: 'JSONB'>, <TokenType.USMALLINT: 'USMALLINT'>, <TokenType.CONSTRAINT: 'CONSTRAINT'>, <TokenType.INTERVAL: 'INTERVAL'>, <TokenType.REFRESH: 'REFRESH'>, <TokenType.STREAMLIT: 'STREAMLIT'>, <TokenType.END: 'END'>, <TokenType.FULL: 'FULL'>, <TokenType.FOREIGN_KEY: 'FOREIGN_KEY'>, <TokenType.DESCRIBE: 'DESCRIBE'>, <TokenType.UINT: 'UINT'>, <TokenType.TDIGEST: 'TDIGEST'>, <TokenType.APPLY: 'APPLY'>, <TokenType.NAME: 'NAME'>, <TokenType.OVERLAPS: 'OVERLAPS'>, <TokenType.TIMESTAMPLTZ: 'TIMESTAMPLTZ'>, <TokenType.TIMETZ: 'TIMETZ'>, <TokenType.PRAGMA: 'PRAGMA'>, <TokenType.ROW: 'ROW'>, <TokenType.ROLLUP: 'ROLLUP'>, <TokenType.INET: 'INET'>, <TokenType.FUNCTION: 'FUNCTION'>, <TokenType.ANY: 'ANY'>, <TokenType.NCHAR: 'NCHAR'>, <TokenType.NVARCHAR: 'NVARCHAR'>, <TokenType.VECTOR: 'VECTOR'>, <TokenType.UDECIMAL: 'UDECIMAL'>, <TokenType.EXECUTE: 'EXECUTE'>, <TokenType.TIMESTAMP_S: 'TIMESTAMP_S'>, <TokenType.TIMESTAMP: 'TIMESTAMP'>, <TokenType.NULL: 'NULL'>, <TokenType.PSEUDO_TYPE: 'PSEUDO_TYPE'>, <TokenType.UPDATE: 'UPDATE'>, <TokenType.TAG: 'TAG'>, <TokenType.IMAGE: 'IMAGE'>, <TokenType.CACHE: 'CACHE'>, <TokenType.SEQUENCE: 'SEQUENCE'>, <TokenType.MEDIUMINT: 'MEDIUMINT'>, <TokenType.SIMPLEAGGREGATEFUNCTION: 'SIMPLEAGGREGATEFUNCTION'>, <TokenType.ISNULL: 'ISNULL'>, <TokenType.DOUBLE: 'DOUBLE'>, <TokenType.AGGREGATEFUNCTION: 'AGGREGATEFUNCTION'>, <TokenType.WAREHOUSE: 'WAREHOUSE'>, <TokenType.CURRENT_TIME: 'CURRENT_TIME'>, <TokenType.OFFSET: 'OFFSET'>}
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
INTERVAL_VARS
ALIAS_TOKENS
ARRAY_CONSTRUCTORS
COMMENT_TABLE_ALIAS_TOKENS
UPDATE_ALIAS_TOKENS
TRIM_TYPES
FUNC_TOKENS
CONJUNCTION
ASSIGNMENT
DISJUNCTION
EQUALITY
COMPARISON
TERM
FACTOR
EXPONENT
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
RANGE_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
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
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):
396    class Generator(generator.Generator):
397        PARAMETER_TOKEN = "$"
398        NAMED_PLACEHOLDER_TOKEN = "$"
399        JOIN_HINTS = False
400        TABLE_HINTS = False
401        QUERY_HINTS = False
402        LIMIT_FETCH = "LIMIT"
403        STRUCT_DELIMITER = ("(", ")")
404        RENAME_TABLE_WITH_DB = False
405        NVL2_SUPPORTED = False
406        SEMI_ANTI_JOIN_WITH_SIDE = False
407        TABLESAMPLE_KEYWORDS = "USING SAMPLE"
408        TABLESAMPLE_SEED_KEYWORD = "REPEATABLE"
409        LAST_DAY_SUPPORTS_DATE_PART = False
410        JSON_KEY_VALUE_PAIR_SEP = ","
411        IGNORE_NULLS_IN_FUNC = True
412        JSON_PATH_BRACKETED_KEY_SUPPORTED = False
413        SUPPORTS_CREATE_TABLE_LIKE = False
414        MULTI_ARG_DISTINCT = False
415        CAN_IMPLEMENT_ARRAY_ANY = True
416        SUPPORTS_TO_NUMBER = False
417        COPY_HAS_INTO_KEYWORD = False
418        STAR_EXCEPT = "EXCLUDE"
419        PAD_FILL_PATTERN_IS_REQUIRED = True
420
421        TRANSFORMS = {
422            **generator.Generator.TRANSFORMS,
423            exp.ApproxDistinct: approx_count_distinct_sql,
424            exp.Array: inline_array_unless_query,
425            exp.ArrayContainsAll: rename_func("ARRAY_HAS_ALL"),
426            exp.ArrayFilter: rename_func("LIST_FILTER"),
427            exp.ArraySize: rename_func("ARRAY_LENGTH"),
428            exp.ArgMax: arg_max_or_min_no_count("ARG_MAX"),
429            exp.ArgMin: arg_max_or_min_no_count("ARG_MIN"),
430            exp.ArraySort: _array_sort_sql,
431            exp.ArraySum: rename_func("LIST_SUM"),
432            exp.BitwiseXor: rename_func("XOR"),
433            exp.CommentColumnConstraint: no_comment_column_constraint_sql,
434            exp.CurrentDate: lambda *_: "CURRENT_DATE",
435            exp.CurrentTime: lambda *_: "CURRENT_TIME",
436            exp.CurrentTimestamp: lambda *_: "CURRENT_TIMESTAMP",
437            exp.DayOfMonth: rename_func("DAYOFMONTH"),
438            exp.DayOfWeek: rename_func("DAYOFWEEK"),
439            exp.DayOfYear: rename_func("DAYOFYEAR"),
440            exp.DataType: _datatype_sql,
441            exp.Date: _date_sql,
442            exp.DateAdd: _date_delta_sql,
443            exp.DateFromParts: rename_func("MAKE_DATE"),
444            exp.DateSub: _date_delta_sql,
445            exp.DateDiff: lambda self, e: self.func(
446                "DATE_DIFF", f"'{e.args.get('unit') or 'DAY'}'", e.expression, e.this
447            ),
448            exp.DateStrToDate: datestrtodate_sql,
449            exp.Datetime: no_datetime_sql,
450            exp.DatetimeSub: _date_delta_sql,
451            exp.DatetimeAdd: _date_delta_sql,
452            exp.DateToDi: lambda self,
453            e: f"CAST(STRFTIME({self.sql(e, 'this')}, {DuckDB.DATEINT_FORMAT}) AS INT)",
454            exp.Decode: lambda self, e: encode_decode_sql(self, e, "DECODE", replace=False),
455            exp.DiToDate: lambda self,
456            e: f"CAST(STRPTIME(CAST({self.sql(e, 'this')} AS TEXT), {DuckDB.DATEINT_FORMAT}) AS DATE)",
457            exp.Encode: lambda self, e: encode_decode_sql(self, e, "ENCODE", replace=False),
458            exp.Explode: rename_func("UNNEST"),
459            exp.IntDiv: lambda self, e: self.binary(e, "//"),
460            exp.IsInf: rename_func("ISINF"),
461            exp.IsNan: rename_func("ISNAN"),
462            exp.JSONExtract: _arrow_json_extract_sql,
463            exp.JSONExtractScalar: _arrow_json_extract_sql,
464            exp.JSONFormat: _json_format_sql,
465            exp.LogicalOr: rename_func("BOOL_OR"),
466            exp.LogicalAnd: rename_func("BOOL_AND"),
467            exp.MD5Digest: lambda self, e: self.func("UNHEX", self.func("MD5", e.this)),
468            exp.MonthsBetween: lambda self, e: self.func(
469                "DATEDIFF",
470                "'month'",
471                exp.cast(e.expression, exp.DataType.Type.TIMESTAMP, copy=True),
472                exp.cast(e.this, exp.DataType.Type.TIMESTAMP, copy=True),
473            ),
474            exp.PercentileCont: rename_func("QUANTILE_CONT"),
475            exp.PercentileDisc: rename_func("QUANTILE_DISC"),
476            # DuckDB doesn't allow qualified columns inside of PIVOT expressions.
477            # See: https://github.com/duckdb/duckdb/blob/671faf92411182f81dce42ac43de8bfb05d9909e/src/planner/binder/tableref/bind_pivot.cpp#L61-L62
478            exp.Pivot: transforms.preprocess([transforms.unqualify_columns]),
479            exp.RegexpExtract: regexp_extract_sql,
480            exp.RegexpReplace: lambda self, e: self.func(
481                "REGEXP_REPLACE",
482                e.this,
483                e.expression,
484                e.args.get("replacement"),
485                e.args.get("modifiers"),
486            ),
487            exp.RegexpLike: rename_func("REGEXP_MATCHES"),
488            exp.RegexpSplit: rename_func("STR_SPLIT_REGEX"),
489            exp.Return: lambda self, e: self.sql(e, "this"),
490            exp.ReturnsProperty: lambda self, e: "TABLE" if isinstance(e.this, exp.Schema) else "",
491            exp.Rand: rename_func("RANDOM"),
492            exp.SafeDivide: no_safe_divide_sql,
493            exp.Split: rename_func("STR_SPLIT"),
494            exp.SortArray: _sort_array_sql,
495            exp.StrPosition: str_position_sql,
496            exp.StrToUnix: lambda self, e: self.func(
497                "EPOCH", self.func("STRPTIME", e.this, self.format_time(e))
498            ),
499            exp.Struct: _struct_sql,
500            exp.TimeAdd: _date_delta_sql,
501            exp.Time: no_time_sql,
502            exp.Timestamp: no_timestamp_sql,
503            exp.TimestampDiff: lambda self, e: self.func(
504                "DATE_DIFF", exp.Literal.string(e.unit), e.expression, e.this
505            ),
506            exp.TimestampTrunc: timestamptrunc_sql(),
507            exp.TimeStrToDate: lambda self, e: self.sql(exp.cast(e.this, exp.DataType.Type.DATE)),
508            exp.TimeStrToTime: timestrtotime_sql,
509            exp.TimeStrToUnix: lambda self, e: self.func(
510                "EPOCH", exp.cast(e.this, exp.DataType.Type.TIMESTAMP)
511            ),
512            exp.TimeToStr: lambda self, e: self.func("STRFTIME", e.this, self.format_time(e)),
513            exp.TimeToUnix: rename_func("EPOCH"),
514            exp.TsOrDiToDi: lambda self,
515            e: f"CAST(SUBSTR(REPLACE(CAST({self.sql(e, 'this')} AS TEXT), '-', ''), 1, 8) AS INT)",
516            exp.TsOrDsAdd: _date_delta_sql,
517            exp.TsOrDsDiff: lambda self, e: self.func(
518                "DATE_DIFF",
519                f"'{e.args.get('unit') or 'DAY'}'",
520                exp.cast(e.expression, exp.DataType.Type.TIMESTAMP),
521                exp.cast(e.this, exp.DataType.Type.TIMESTAMP),
522            ),
523            exp.UnixToStr: lambda self, e: self.func(
524                "STRFTIME", self.func("TO_TIMESTAMP", e.this), self.format_time(e)
525            ),
526            exp.UnixToTime: _unix_to_time_sql,
527            exp.UnixToTimeStr: lambda self, e: f"CAST(TO_TIMESTAMP({self.sql(e, 'this')}) AS TEXT)",
528            exp.VariancePop: rename_func("VAR_POP"),
529            exp.WeekOfYear: rename_func("WEEKOFYEAR"),
530            exp.Xor: bool_xor_sql,
531        }
532
533        SUPPORTED_JSON_PATH_PARTS = {
534            exp.JSONPathKey,
535            exp.JSONPathRoot,
536            exp.JSONPathSubscript,
537            exp.JSONPathWildcard,
538        }
539
540        TYPE_MAPPING = {
541            **generator.Generator.TYPE_MAPPING,
542            exp.DataType.Type.BINARY: "BLOB",
543            exp.DataType.Type.BPCHAR: "TEXT",
544            exp.DataType.Type.CHAR: "TEXT",
545            exp.DataType.Type.FLOAT: "REAL",
546            exp.DataType.Type.NCHAR: "TEXT",
547            exp.DataType.Type.NVARCHAR: "TEXT",
548            exp.DataType.Type.UINT: "UINTEGER",
549            exp.DataType.Type.VARBINARY: "BLOB",
550            exp.DataType.Type.ROWVERSION: "BLOB",
551            exp.DataType.Type.VARCHAR: "TEXT",
552            exp.DataType.Type.TIMESTAMPNTZ: "TIMESTAMP",
553            exp.DataType.Type.TIMESTAMP_S: "TIMESTAMP_S",
554            exp.DataType.Type.TIMESTAMP_MS: "TIMESTAMP_MS",
555            exp.DataType.Type.TIMESTAMP_NS: "TIMESTAMP_NS",
556        }
557
558        # https://github.com/duckdb/duckdb/blob/ff7f24fd8e3128d94371827523dae85ebaf58713/third_party/libpg_query/grammar/keywords/reserved_keywords.list#L1-L77
559        RESERVED_KEYWORDS = {
560            "array",
561            "analyse",
562            "union",
563            "all",
564            "when",
565            "in_p",
566            "default",
567            "create_p",
568            "window",
569            "asymmetric",
570            "to",
571            "else",
572            "localtime",
573            "from",
574            "end_p",
575            "select",
576            "current_date",
577            "foreign",
578            "with",
579            "grant",
580            "session_user",
581            "or",
582            "except",
583            "references",
584            "fetch",
585            "limit",
586            "group_p",
587            "leading",
588            "into",
589            "collate",
590            "offset",
591            "do",
592            "then",
593            "localtimestamp",
594            "check_p",
595            "lateral_p",
596            "current_role",
597            "where",
598            "asc_p",
599            "placing",
600            "desc_p",
601            "user",
602            "unique",
603            "initially",
604            "column",
605            "both",
606            "some",
607            "as",
608            "any",
609            "only",
610            "deferrable",
611            "null_p",
612            "current_time",
613            "true_p",
614            "table",
615            "case",
616            "trailing",
617            "variadic",
618            "for",
619            "on",
620            "distinct",
621            "false_p",
622            "not",
623            "constraint",
624            "current_timestamp",
625            "returning",
626            "primary",
627            "intersect",
628            "having",
629            "analyze",
630            "current_user",
631            "and",
632            "cast",
633            "symmetric",
634            "using",
635            "order",
636            "current_catalog",
637        }
638
639        UNWRAPPED_INTERVAL_VALUES = (exp.Literal, exp.Paren)
640
641        # DuckDB doesn't generally support CREATE TABLE .. properties
642        # https://duckdb.org/docs/sql/statements/create_table.html
643        PROPERTIES_LOCATION = {
644            prop: exp.Properties.Location.UNSUPPORTED
645            for prop in generator.Generator.PROPERTIES_LOCATION
646        }
647
648        # There are a few exceptions (e.g. temporary tables) which are supported or
649        # can be transpiled to DuckDB, so we explicitly override them accordingly
650        PROPERTIES_LOCATION[exp.LikeProperty] = exp.Properties.Location.POST_SCHEMA
651        PROPERTIES_LOCATION[exp.TemporaryProperty] = exp.Properties.Location.POST_CREATE
652        PROPERTIES_LOCATION[exp.ReturnsProperty] = exp.Properties.Location.POST_ALIAS
653
654        def strtotime_sql(self, expression: exp.StrToTime) -> str:
655            if expression.args.get("safe"):
656                formatted_time = self.format_time(expression)
657                return f"CAST({self.func('TRY_STRPTIME', expression.this, formatted_time)} AS TIMESTAMP)"
658            return str_to_time_sql(self, expression)
659
660        def strtodate_sql(self, expression: exp.StrToDate) -> str:
661            if expression.args.get("safe"):
662                formatted_time = self.format_time(expression)
663                return f"CAST({self.func('TRY_STRPTIME', expression.this, formatted_time)} AS DATE)"
664            return f"CAST({str_to_time_sql(self, expression)} AS DATE)"
665
666        def parsejson_sql(self, expression: exp.ParseJSON) -> str:
667            arg = expression.this
668            if expression.args.get("safe"):
669                return self.sql(exp.case().when(exp.func("json_valid", arg), arg).else_(exp.null()))
670            return self.func("JSON", arg)
671
672        def timefromparts_sql(self, expression: exp.TimeFromParts) -> str:
673            nano = expression.args.get("nano")
674            if nano is not None:
675                expression.set(
676                    "sec", expression.args["sec"] + nano.pop() / exp.Literal.number(1000000000.0)
677                )
678
679            return rename_func("MAKE_TIME")(self, expression)
680
681        def timestampfromparts_sql(self, expression: exp.TimestampFromParts) -> str:
682            sec = expression.args["sec"]
683
684            milli = expression.args.get("milli")
685            if milli is not None:
686                sec += milli.pop() / exp.Literal.number(1000.0)
687
688            nano = expression.args.get("nano")
689            if nano is not None:
690                sec += nano.pop() / exp.Literal.number(1000000000.0)
691
692            if milli or nano:
693                expression.set("sec", sec)
694
695            return rename_func("MAKE_TIMESTAMP")(self, expression)
696
697        def tablesample_sql(
698            self,
699            expression: exp.TableSample,
700            sep: str = " AS ",
701            tablesample_keyword: t.Optional[str] = None,
702        ) -> str:
703            if not isinstance(expression.parent, exp.Select):
704                # This sample clause only applies to a single source, not the entire resulting relation
705                tablesample_keyword = "TABLESAMPLE"
706
707            if expression.args.get("size"):
708                method = expression.args.get("method")
709                if method and method.name.upper() != "RESERVOIR":
710                    self.unsupported(
711                        f"Sampling method {method} is not supported with a discrete sample count, "
712                        "defaulting to reservoir sampling"
713                    )
714                    expression.set("method", exp.var("RESERVOIR"))
715
716            return super().tablesample_sql(
717                expression, sep=sep, tablesample_keyword=tablesample_keyword
718            )
719
720        def interval_sql(self, expression: exp.Interval) -> str:
721            multiplier: t.Optional[int] = None
722            unit = expression.text("unit").lower()
723
724            if unit.startswith("week"):
725                multiplier = 7
726            if unit.startswith("quarter"):
727                multiplier = 90
728
729            if multiplier:
730                return f"({multiplier} * {super().interval_sql(exp.Interval(this=expression.this, unit=exp.var('DAY')))})"
731
732            return super().interval_sql(expression)
733
734        def columndef_sql(self, expression: exp.ColumnDef, sep: str = " ") -> str:
735            if isinstance(expression.parent, exp.UserDefinedFunction):
736                return self.sql(expression, "this")
737            return super().columndef_sql(expression, sep)
738
739        def join_sql(self, expression: exp.Join) -> str:
740            if (
741                expression.side == "LEFT"
742                and not expression.args.get("on")
743                and isinstance(expression.this, exp.Unnest)
744            ):
745                # Some dialects support `LEFT JOIN UNNEST(...)` without an explicit ON clause
746                # DuckDB doesn't, but we can just add a dummy ON clause that is always true
747                return super().join_sql(expression.on(exp.true()))
748
749            return super().join_sql(expression)
750
751        def generateseries_sql(self, expression: exp.GenerateSeries) -> str:
752            # GENERATE_SERIES(a, b) -> [a, b], RANGE(a, b) -> [a, b)
753            if expression.args.get("is_end_exclusive"):
754                return rename_func("RANGE")(self, expression)
755
756            return self.function_fallback_sql(expression)
757
758        def bracket_sql(self, expression: exp.Bracket) -> str:
759            this = expression.this
760            if isinstance(this, exp.Array):
761                this.replace(exp.paren(this))
762
763            bracket = super().bracket_sql(expression)
764
765            if not expression.args.get("returns_list_for_maps"):
766                if not this.type:
767                    from sqlglot.optimizer.annotate_types import annotate_types
768
769                    this = annotate_types(this)
770
771                if this.is_type(exp.DataType.Type.MAP):
772                    bracket = f"({bracket})[1]"
773
774            return bracket
775
776        def withingroup_sql(self, expression: exp.WithinGroup) -> str:
777            expression_sql = self.sql(expression, "expression")
778
779            func = expression.this
780            if isinstance(func, exp.PERCENTILES):
781                # Make the order key the first arg and slide the fraction to the right
782                # https://duckdb.org/docs/sql/aggregates#ordered-set-aggregate-functions
783                order_col = expression.find(exp.Ordered)
784                if order_col:
785                    func.set("expression", func.this)
786                    func.set("this", order_col.this)
787
788            this = self.sql(expression, "this").rstrip(")")
789
790            return f"{this}{expression_sql})"
791
792        def length_sql(self, expression: exp.Length) -> str:
793            arg = expression.this
794
795            # Dialects like BQ and Snowflake also accept binary values as args, so
796            # DDB will attempt to infer the type or resort to case/when resolution
797            if not expression.args.get("binary") or arg.is_string:
798                return self.func("LENGTH", arg)
799
800            if not arg.type:
801                from sqlglot.optimizer.annotate_types import annotate_types
802
803                arg = annotate_types(arg)
804
805            if arg.is_type(*exp.DataType.TEXT_TYPES):
806                return self.func("LENGTH", arg)
807
808            # We need these casts to make duckdb's static type checker happy
809            blob = exp.cast(arg, exp.DataType.Type.VARBINARY)
810            varchar = exp.cast(arg, exp.DataType.Type.VARCHAR)
811
812            case = (
813                exp.case(self.func("TYPEOF", arg))
814                .when(
815                    "'VARCHAR'", exp.Anonymous(this="LENGTH", expressions=[varchar])
816                )  # anonymous to break length_sql recursion
817                .when("'BLOB'", self.func("OCTET_LENGTH", blob))
818            )
819
820            return self.sql(case)
821
822        def objectinsert_sql(self, expression: exp.ObjectInsert) -> str:
823            this = expression.this
824            key = expression.args.get("key")
825            key_sql = key.name if isinstance(key, exp.Expression) else ""
826            value_sql = self.sql(expression, "value")
827
828            kv_sql = f"{key_sql} := {value_sql}"
829
830            # If the input struct is empty e.g. transpiling OBJECT_INSERT(OBJECT_CONSTRUCT(), key, value) from Snowflake
831            # then we can generate STRUCT_PACK which will build it since STRUCT_INSERT({}, key := value) is not valid DuckDB
832            if isinstance(this, exp.Struct) and not this.expressions:
833                return self.func("STRUCT_PACK", kv_sql)
834
835            return self.func("STRUCT_INSERT", this, kv_sql)

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
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.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.EncodeColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.EphemeralColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ExcludeColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ExecuteAsProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ExternalProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.GlobalProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.HeapProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.IcebergProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.InheritsProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.InlineLengthColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.InputModelProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.IntervalSpan'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.LanguageProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.LocationProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.LogProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.MaterializedProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.NonClusteredColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.NoPrimaryIndexProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.NotForReplicationColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.OnCommitProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.OnProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.OnUpdateColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.OutputModelProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.PathColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ProjectionPolicyColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.RemoteWithConnectionModelProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ReturnsProperty'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.SampleProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SecureProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SetConfigProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SetProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SettingsProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SharingProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SqlReadWriteProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SqlSecurityProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.StabilityProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.StrictProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.TemporaryProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.TagColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.TitleColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ToMap'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ToTableProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.TransformModelProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.TransientProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.UppercaseColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.UnloggedProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.VarMap'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ViewAttributeProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.VolatileProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.WithJournalTableProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.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.ArrayContainsAll'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.ArrayFilter'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.ArraySize'>: <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.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 DuckDB.Generator.<lambda>>, <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.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.JSONExtract'>: <function _arrow_json_extract_sql>, <class 'sqlglot.expressions.JSONExtractScalar'>: <function _arrow_json_extract_sql>, <class 'sqlglot.expressions.JSONFormat'>: <function _json_format_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.RegexpExtract'>: <function regexp_extract_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.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.TimeAdd'>: <function _date_delta_sql>, <class 'sqlglot.expressions.Time'>: <function no_time_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.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>}
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 = {'references', 'grant', 'to', 'check_p', 'current_date', 'variadic', 'from', 'user', 'localtime', 'order', 'current_time', 'true_p', 'select', 'as', 'except', 'asymmetric', 'group_p', 'do', 'offset', 'unique', 'deferrable', 'else', 'create_p', 'primary', 'limit', 'when', 'case', 'default', 'current_catalog', 'column', 'not', 'localtimestamp', 'null_p', 'current_role', 'symmetric', 'trailing', 'analyse', 'only', 'foreign', 'both', 'using', 'collate', 'in_p', 'intersect', 'analyze', 'false_p', 'window', 'leading', 'all', 'into', 'for', 'desc_p', 'having', 'and', 'constraint', 'returning', 'asc_p', 'current_user', 'distinct', 'lateral_p', 'or', 'initially', 'on', 'placing', 'union', 'table', 'array', 'fetch', 'with', 'cast', 'end_p', 'current_timestamp', 'some', 'any', 'where', 'session_user', 'then'}
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.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.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.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.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.WithSchemaBindingProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.WithSystemVersioningProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>}
def strtotime_sql(self, expression: sqlglot.expressions.StrToTime) -> str:
654        def strtotime_sql(self, expression: exp.StrToTime) -> str:
655            if expression.args.get("safe"):
656                formatted_time = self.format_time(expression)
657                return f"CAST({self.func('TRY_STRPTIME', expression.this, formatted_time)} AS TIMESTAMP)"
658            return str_to_time_sql(self, expression)
def strtodate_sql(self, expression: sqlglot.expressions.StrToDate) -> str:
660        def strtodate_sql(self, expression: exp.StrToDate) -> str:
661            if expression.args.get("safe"):
662                formatted_time = self.format_time(expression)
663                return f"CAST({self.func('TRY_STRPTIME', expression.this, formatted_time)} AS DATE)"
664            return f"CAST({str_to_time_sql(self, expression)} AS DATE)"
def parsejson_sql(self, expression: sqlglot.expressions.ParseJSON) -> str:
666        def parsejson_sql(self, expression: exp.ParseJSON) -> str:
667            arg = expression.this
668            if expression.args.get("safe"):
669                return self.sql(exp.case().when(exp.func("json_valid", arg), arg).else_(exp.null()))
670            return self.func("JSON", arg)
def timefromparts_sql(self, expression: sqlglot.expressions.TimeFromParts) -> str:
672        def timefromparts_sql(self, expression: exp.TimeFromParts) -> str:
673            nano = expression.args.get("nano")
674            if nano is not None:
675                expression.set(
676                    "sec", expression.args["sec"] + nano.pop() / exp.Literal.number(1000000000.0)
677                )
678
679            return rename_func("MAKE_TIME")(self, expression)
def timestampfromparts_sql(self, expression: sqlglot.expressions.TimestampFromParts) -> str:
681        def timestampfromparts_sql(self, expression: exp.TimestampFromParts) -> str:
682            sec = expression.args["sec"]
683
684            milli = expression.args.get("milli")
685            if milli is not None:
686                sec += milli.pop() / exp.Literal.number(1000.0)
687
688            nano = expression.args.get("nano")
689            if nano is not None:
690                sec += nano.pop() / exp.Literal.number(1000000000.0)
691
692            if milli or nano:
693                expression.set("sec", sec)
694
695            return rename_func("MAKE_TIMESTAMP")(self, expression)
def tablesample_sql( self, expression: sqlglot.expressions.TableSample, sep: str = ' AS ', tablesample_keyword: Optional[str] = None) -> str:
697        def tablesample_sql(
698            self,
699            expression: exp.TableSample,
700            sep: str = " AS ",
701            tablesample_keyword: t.Optional[str] = None,
702        ) -> str:
703            if not isinstance(expression.parent, exp.Select):
704                # This sample clause only applies to a single source, not the entire resulting relation
705                tablesample_keyword = "TABLESAMPLE"
706
707            if expression.args.get("size"):
708                method = expression.args.get("method")
709                if method and method.name.upper() != "RESERVOIR":
710                    self.unsupported(
711                        f"Sampling method {method} is not supported with a discrete sample count, "
712                        "defaulting to reservoir sampling"
713                    )
714                    expression.set("method", exp.var("RESERVOIR"))
715
716            return super().tablesample_sql(
717                expression, sep=sep, tablesample_keyword=tablesample_keyword
718            )
def interval_sql(self, expression: sqlglot.expressions.Interval) -> str:
720        def interval_sql(self, expression: exp.Interval) -> str:
721            multiplier: t.Optional[int] = None
722            unit = expression.text("unit").lower()
723
724            if unit.startswith("week"):
725                multiplier = 7
726            if unit.startswith("quarter"):
727                multiplier = 90
728
729            if multiplier:
730                return f"({multiplier} * {super().interval_sql(exp.Interval(this=expression.this, unit=exp.var('DAY')))})"
731
732            return super().interval_sql(expression)
def columndef_sql(self, expression: sqlglot.expressions.ColumnDef, sep: str = ' ') -> str:
734        def columndef_sql(self, expression: exp.ColumnDef, sep: str = " ") -> str:
735            if isinstance(expression.parent, exp.UserDefinedFunction):
736                return self.sql(expression, "this")
737            return super().columndef_sql(expression, sep)
def join_sql(self, expression: sqlglot.expressions.Join) -> str:
739        def join_sql(self, expression: exp.Join) -> str:
740            if (
741                expression.side == "LEFT"
742                and not expression.args.get("on")
743                and isinstance(expression.this, exp.Unnest)
744            ):
745                # Some dialects support `LEFT JOIN UNNEST(...)` without an explicit ON clause
746                # DuckDB doesn't, but we can just add a dummy ON clause that is always true
747                return super().join_sql(expression.on(exp.true()))
748
749            return super().join_sql(expression)
def generateseries_sql(self, expression: sqlglot.expressions.GenerateSeries) -> str:
751        def generateseries_sql(self, expression: exp.GenerateSeries) -> str:
752            # GENERATE_SERIES(a, b) -> [a, b], RANGE(a, b) -> [a, b)
753            if expression.args.get("is_end_exclusive"):
754                return rename_func("RANGE")(self, expression)
755
756            return self.function_fallback_sql(expression)
def bracket_sql(self, expression: sqlglot.expressions.Bracket) -> str:
758        def bracket_sql(self, expression: exp.Bracket) -> str:
759            this = expression.this
760            if isinstance(this, exp.Array):
761                this.replace(exp.paren(this))
762
763            bracket = super().bracket_sql(expression)
764
765            if not expression.args.get("returns_list_for_maps"):
766                if not this.type:
767                    from sqlglot.optimizer.annotate_types import annotate_types
768
769                    this = annotate_types(this)
770
771                if this.is_type(exp.DataType.Type.MAP):
772                    bracket = f"({bracket})[1]"
773
774            return bracket
def withingroup_sql(self, expression: sqlglot.expressions.WithinGroup) -> str:
776        def withingroup_sql(self, expression: exp.WithinGroup) -> str:
777            expression_sql = self.sql(expression, "expression")
778
779            func = expression.this
780            if isinstance(func, exp.PERCENTILES):
781                # Make the order key the first arg and slide the fraction to the right
782                # https://duckdb.org/docs/sql/aggregates#ordered-set-aggregate-functions
783                order_col = expression.find(exp.Ordered)
784                if order_col:
785                    func.set("expression", func.this)
786                    func.set("this", order_col.this)
787
788            this = self.sql(expression, "this").rstrip(")")
789
790            return f"{this}{expression_sql})"
def length_sql(self, expression: sqlglot.expressions.Length) -> str:
792        def length_sql(self, expression: exp.Length) -> str:
793            arg = expression.this
794
795            # Dialects like BQ and Snowflake also accept binary values as args, so
796            # DDB will attempt to infer the type or resort to case/when resolution
797            if not expression.args.get("binary") or arg.is_string:
798                return self.func("LENGTH", arg)
799
800            if not arg.type:
801                from sqlglot.optimizer.annotate_types import annotate_types
802
803                arg = annotate_types(arg)
804
805            if arg.is_type(*exp.DataType.TEXT_TYPES):
806                return self.func("LENGTH", arg)
807
808            # We need these casts to make duckdb's static type checker happy
809            blob = exp.cast(arg, exp.DataType.Type.VARBINARY)
810            varchar = exp.cast(arg, exp.DataType.Type.VARCHAR)
811
812            case = (
813                exp.case(self.func("TYPEOF", arg))
814                .when(
815                    "'VARCHAR'", exp.Anonymous(this="LENGTH", expressions=[varchar])
816                )  # anonymous to break length_sql recursion
817                .when("'BLOB'", self.func("OCTET_LENGTH", blob))
818            )
819
820            return self.sql(case)
def objectinsert_sql(self, expression: sqlglot.expressions.ObjectInsert) -> str:
822        def objectinsert_sql(self, expression: exp.ObjectInsert) -> str:
823            this = expression.this
824            key = expression.args.get("key")
825            key_sql = key.name if isinstance(key, exp.Expression) else ""
826            value_sql = self.sql(expression, "value")
827
828            kv_sql = f"{key_sql} := {value_sql}"
829
830            # If the input struct is empty e.g. transpiling OBJECT_INSERT(OBJECT_CONSTRUCT(), key, value) from Snowflake
831            # then we can generate STRUCT_PACK which will build it since STRUCT_INSERT({}, key := value) is not valid DuckDB
832            if isinstance(this, exp.Struct) and not this.expressions:
833                return self.func("STRUCT_PACK", kv_sql)
834
835            return self.func("STRUCT_INSERT", this, kv_sql)
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
EXPLICIT_SET_OP
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
PARSE_JSON_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
except_sql
except_op
fetch_sql
filter_sql
hint_sql
indexparameters_sql
index_sql
identifier_sql
hex_sql
lowerhex_sql
inputoutputformat_sql
national_sql
partition_sql
properties_sql
root_properties
properties
with_properties
locate_properties
property_name
property_sql
likeproperty_sql
fallbackproperty_sql
journalproperty_sql
freespaceproperty_sql
checksumproperty_sql
mergeblockratioproperty_sql
datablocksizeproperty_sql
blockcompressionproperty_sql
isolatedloadingproperty_sql
partitionboundspec_sql
partitionedofproperty_sql
lockingproperty_sql
withdataproperty_sql
withsystemversioningproperty_sql
insert_sql
intersect_sql
intersect_op
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
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
set_operations
union_sql
union_op
unnest_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
renametable_sql
renamecolumn_sql
alterset_sql
altertable_sql
add_column_sql
droppartition_sql
addconstraint_sql
distinct_sql
ignorenulls_sql
respectnulls_sql
havingmax_sql
intdiv_sql
dpipe_sql
div_sql
overlaps_sql
distance_sql
dot_sql
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
oncluster_sql
clusteredbyproperty_sql
anyvalue_sql
querytransform_sql
indexconstraintoption_sql
checkcolumnconstraint_sql
indexcolumnconstraint_sql
nvl2_sql
comprehension_sql
columnprefix_sql
opclass_sql
predict_sql
forin_sql
refresh_sql
operator_sql
toarray_sql
tsordstotime_sql
tsordstotimestamp_sql
tsordstodate_sql
unixdate_sql
lastday_sql
dateadd_sql
arrayany_sql
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