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.dialects.dialect import (
  7    Dialect,
  8    NormalizationStrategy,
  9    approx_count_distinct_sql,
 10    arg_max_or_min_no_count,
 11    arrow_json_extract_sql,
 12    binary_from_function,
 13    bool_xor_sql,
 14    date_trunc_to_time,
 15    datestrtodate_sql,
 16    encode_decode_sql,
 17    build_formatted_time,
 18    inline_array_unless_query,
 19    no_comment_column_constraint_sql,
 20    no_safe_divide_sql,
 21    no_timestamp_sql,
 22    pivot_column_names,
 23    regexp_extract_sql,
 24    rename_func,
 25    str_position_sql,
 26    str_to_time_sql,
 27    timestamptrunc_sql,
 28    timestrtotime_sql,
 29    unit_to_var,
 30)
 31from sqlglot.helper import flatten, seq_get
 32from sqlglot.tokens import TokenType
 33
 34
 35def _ts_or_ds_add_sql(self: DuckDB.Generator, expression: exp.TsOrDsAdd) -> str:
 36    this = self.sql(expression, "this")
 37    interval = self.sql(exp.Interval(this=expression.expression, unit=unit_to_var(expression)))
 38    return f"CAST({this} AS {self.sql(expression.return_type)}) + {interval}"
 39
 40
 41def _date_delta_sql(
 42    self: DuckDB.Generator, expression: exp.DateAdd | exp.DateSub | exp.TimeAdd
 43) -> str:
 44    this = self.sql(expression, "this")
 45    unit = unit_to_var(expression)
 46    op = "+" if isinstance(expression, (exp.DateAdd, exp.TimeAdd)) else "-"
 47    return f"{this} {op} {self.sql(exp.Interval(this=expression.expression, unit=unit))}"
 48
 49
 50# BigQuery -> DuckDB conversion for the DATE function
 51def _date_sql(self: DuckDB.Generator, expression: exp.Date) -> str:
 52    result = f"CAST({self.sql(expression, 'this')} AS DATE)"
 53    zone = self.sql(expression, "zone")
 54
 55    if zone:
 56        date_str = self.func("STRFTIME", result, "'%d/%m/%Y'")
 57        date_str = f"{date_str} || ' ' || {zone}"
 58
 59        # This will create a TIMESTAMP with time zone information
 60        result = self.func("STRPTIME", date_str, "'%d/%m/%Y %Z'")
 61
 62    return result
 63
 64
 65def _array_sort_sql(self: DuckDB.Generator, expression: exp.ArraySort) -> str:
 66    if expression.expression:
 67        self.unsupported("DuckDB ARRAY_SORT does not support a comparator")
 68    return self.func("ARRAY_SORT", expression.this)
 69
 70
 71def _sort_array_sql(self: DuckDB.Generator, expression: exp.SortArray) -> str:
 72    name = "ARRAY_REVERSE_SORT" if expression.args.get("asc") == exp.false() else "ARRAY_SORT"
 73    return self.func(name, expression.this)
 74
 75
 76def _build_sort_array_desc(args: t.List) -> exp.Expression:
 77    return exp.SortArray(this=seq_get(args, 0), asc=exp.false())
 78
 79
 80def _build_date_diff(args: t.List) -> exp.Expression:
 81    return exp.DateDiff(this=seq_get(args, 2), expression=seq_get(args, 1), unit=seq_get(args, 0))
 82
 83
 84def _build_generate_series(end_exclusive: bool = False) -> t.Callable[[t.List], exp.GenerateSeries]:
 85    def _builder(args: t.List) -> exp.GenerateSeries:
 86        # Check https://duckdb.org/docs/sql/functions/nested.html#range-functions
 87        if len(args) == 1:
 88            # DuckDB uses 0 as a default for the series' start when it's omitted
 89            args.insert(0, exp.Literal.number("0"))
 90
 91        gen_series = exp.GenerateSeries.from_arg_list(args)
 92        gen_series.set("is_end_exclusive", end_exclusive)
 93
 94        return gen_series
 95
 96    return _builder
 97
 98
 99def _build_make_timestamp(args: t.List) -> exp.Expression:
100    if len(args) == 1:
101        return exp.UnixToTime(this=seq_get(args, 0), scale=exp.UnixToTime.MICROS)
102
103    return exp.TimestampFromParts(
104        year=seq_get(args, 0),
105        month=seq_get(args, 1),
106        day=seq_get(args, 2),
107        hour=seq_get(args, 3),
108        min=seq_get(args, 4),
109        sec=seq_get(args, 5),
110    )
111
112
113def _struct_sql(self: DuckDB.Generator, expression: exp.Struct) -> str:
114    args: t.List[str] = []
115    for i, expr in enumerate(expression.expressions):
116        if isinstance(expr, exp.PropertyEQ):
117            key = expr.name
118            value = expr.expression
119        else:
120            key = f"_{i}"
121            value = expr
122
123        args.append(f"{self.sql(exp.Literal.string(key))}: {self.sql(value)}")
124
125    return f"{{{', '.join(args)}}}"
126
127
128def _datatype_sql(self: DuckDB.Generator, expression: exp.DataType) -> str:
129    if expression.is_type("array"):
130        return f"{self.expressions(expression, flat=True)}[]"
131
132    # Type TIMESTAMP / TIME WITH TIME ZONE does not support any modifiers
133    if expression.is_type("timestamptz", "timetz"):
134        return expression.this.value
135
136    return self.datatype_sql(expression)
137
138
139def _json_format_sql(self: DuckDB.Generator, expression: exp.JSONFormat) -> str:
140    sql = self.func("TO_JSON", expression.this, expression.args.get("options"))
141    return f"CAST({sql} AS TEXT)"
142
143
144def _unix_to_time_sql(self: DuckDB.Generator, expression: exp.UnixToTime) -> str:
145    scale = expression.args.get("scale")
146    timestamp = expression.this
147
148    if scale in (None, exp.UnixToTime.SECONDS):
149        return self.func("TO_TIMESTAMP", timestamp)
150    if scale == exp.UnixToTime.MILLIS:
151        return self.func("EPOCH_MS", timestamp)
152    if scale == exp.UnixToTime.MICROS:
153        return self.func("MAKE_TIMESTAMP", timestamp)
154
155    return self.func("TO_TIMESTAMP", exp.Div(this=timestamp, expression=exp.func("POW", 10, scale)))
156
157
158def _rename_unless_within_group(
159    a: str, b: str
160) -> t.Callable[[DuckDB.Generator, exp.Expression], str]:
161    return lambda self, expression: (
162        self.func(a, *flatten(expression.args.values()))
163        if isinstance(expression.find_ancestor(exp.Select, exp.WithinGroup), exp.WithinGroup)
164        else self.func(b, *flatten(expression.args.values()))
165    )
166
167
168class DuckDB(Dialect):
169    NULL_ORDERING = "nulls_are_last"
170    SUPPORTS_USER_DEFINED_TYPES = False
171    SAFE_DIVISION = True
172    INDEX_OFFSET = 1
173    CONCAT_COALESCE = True
174
175    # https://duckdb.org/docs/sql/introduction.html#creating-a-new-table
176    NORMALIZATION_STRATEGY = NormalizationStrategy.CASE_INSENSITIVE
177
178    def to_json_path(self, path: t.Optional[exp.Expression]) -> t.Optional[exp.Expression]:
179        if isinstance(path, exp.Literal):
180            # DuckDB also supports the JSON pointer syntax, where every path starts with a `/`.
181            # Additionally, it allows accessing the back of lists using the `[#-i]` syntax.
182            # This check ensures we'll avoid trying to parse these as JSON paths, which can
183            # either result in a noisy warning or in an invalid representation of the path.
184            path_text = path.name
185            if path_text.startswith("/") or "[#" in path_text:
186                return path
187
188        return super().to_json_path(path)
189
190    class Tokenizer(tokens.Tokenizer):
191        HEREDOC_STRINGS = ["$"]
192
193        HEREDOC_TAG_IS_IDENTIFIER = True
194        HEREDOC_STRING_ALTERNATIVE = TokenType.PARAMETER
195
196        KEYWORDS = {
197            **tokens.Tokenizer.KEYWORDS,
198            "//": TokenType.DIV,
199            "ATTACH": TokenType.COMMAND,
200            "BINARY": TokenType.VARBINARY,
201            "BITSTRING": TokenType.BIT,
202            "BPCHAR": TokenType.TEXT,
203            "CHAR": TokenType.TEXT,
204            "CHARACTER VARYING": TokenType.TEXT,
205            "EXCLUDE": TokenType.EXCEPT,
206            "LOGICAL": TokenType.BOOLEAN,
207            "ONLY": TokenType.ONLY,
208            "PIVOT_WIDER": TokenType.PIVOT,
209            "POSITIONAL": TokenType.POSITIONAL,
210            "SIGNED": TokenType.INT,
211            "STRING": TokenType.VARCHAR,
212            "UBIGINT": TokenType.UBIGINT,
213            "UINTEGER": TokenType.UINT,
214            "USMALLINT": TokenType.USMALLINT,
215            "UTINYINT": TokenType.UTINYINT,
216            "TIMESTAMP_S": TokenType.TIMESTAMP_S,
217            "TIMESTAMP_MS": TokenType.TIMESTAMP_MS,
218            "TIMESTAMP_NS": TokenType.TIMESTAMP_NS,
219            "TIMESTAMP_US": TokenType.TIMESTAMP,
220        }
221
222        SINGLE_TOKENS = {
223            **tokens.Tokenizer.SINGLE_TOKENS,
224            "$": TokenType.PARAMETER,
225        }
226
227    class Parser(parser.Parser):
228        BITWISE = {
229            **parser.Parser.BITWISE,
230            TokenType.TILDA: exp.RegexpLike,
231        }
232
233        FUNCTIONS_WITH_ALIASED_ARGS = {*parser.Parser.FUNCTIONS_WITH_ALIASED_ARGS, "STRUCT_PACK"}
234
235        FUNCTIONS = {
236            **parser.Parser.FUNCTIONS,
237            "ARRAY_HAS": exp.ArrayContains.from_arg_list,
238            "ARRAY_REVERSE_SORT": _build_sort_array_desc,
239            "ARRAY_SORT": exp.SortArray.from_arg_list,
240            "DATEDIFF": _build_date_diff,
241            "DATE_DIFF": _build_date_diff,
242            "DATE_TRUNC": date_trunc_to_time,
243            "DATETRUNC": date_trunc_to_time,
244            "DECODE": lambda args: exp.Decode(
245                this=seq_get(args, 0), charset=exp.Literal.string("utf-8")
246            ),
247            "ENCODE": lambda args: exp.Encode(
248                this=seq_get(args, 0), charset=exp.Literal.string("utf-8")
249            ),
250            "EPOCH": exp.TimeToUnix.from_arg_list,
251            "EPOCH_MS": lambda args: exp.UnixToTime(
252                this=seq_get(args, 0), scale=exp.UnixToTime.MILLIS
253            ),
254            "JSON": exp.ParseJSON.from_arg_list,
255            "JSON_EXTRACT_PATH": parser.build_extract_json_with_path(exp.JSONExtract),
256            "JSON_EXTRACT_STRING": parser.build_extract_json_with_path(exp.JSONExtractScalar),
257            "LIST_HAS": exp.ArrayContains.from_arg_list,
258            "LIST_REVERSE_SORT": _build_sort_array_desc,
259            "LIST_SORT": exp.SortArray.from_arg_list,
260            "LIST_VALUE": exp.Array.from_arg_list,
261            "MAKE_TIME": exp.TimeFromParts.from_arg_list,
262            "MAKE_TIMESTAMP": _build_make_timestamp,
263            "MEDIAN": lambda args: exp.PercentileCont(
264                this=seq_get(args, 0), expression=exp.Literal.number(0.5)
265            ),
266            "QUANTILE_CONT": exp.PercentileCont.from_arg_list,
267            "QUANTILE_DISC": exp.PercentileDisc.from_arg_list,
268            "REGEXP_EXTRACT": lambda args: exp.RegexpExtract(
269                this=seq_get(args, 0), expression=seq_get(args, 1), group=seq_get(args, 2)
270            ),
271            "REGEXP_MATCHES": exp.RegexpLike.from_arg_list,
272            "REGEXP_REPLACE": lambda args: exp.RegexpReplace(
273                this=seq_get(args, 0),
274                expression=seq_get(args, 1),
275                replacement=seq_get(args, 2),
276                modifiers=seq_get(args, 3),
277            ),
278            "STRFTIME": build_formatted_time(exp.TimeToStr, "duckdb"),
279            "STRING_SPLIT": exp.Split.from_arg_list,
280            "STRING_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
281            "STRING_TO_ARRAY": exp.Split.from_arg_list,
282            "STRPTIME": build_formatted_time(exp.StrToTime, "duckdb"),
283            "STRUCT_PACK": exp.Struct.from_arg_list,
284            "STR_SPLIT": exp.Split.from_arg_list,
285            "STR_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
286            "TO_TIMESTAMP": exp.UnixToTime.from_arg_list,
287            "UNNEST": exp.Explode.from_arg_list,
288            "XOR": binary_from_function(exp.BitwiseXor),
289            "GENERATE_SERIES": _build_generate_series(),
290            "RANGE": _build_generate_series(end_exclusive=True),
291        }
292
293        FUNCTION_PARSERS = parser.Parser.FUNCTION_PARSERS.copy()
294        FUNCTION_PARSERS.pop("DECODE")
295
296        NO_PAREN_FUNCTION_PARSERS = {
297            **parser.Parser.NO_PAREN_FUNCTION_PARSERS,
298            "MAP": lambda self: self._parse_map(),
299        }
300
301        TABLE_ALIAS_TOKENS = parser.Parser.TABLE_ALIAS_TOKENS - {
302            TokenType.SEMI,
303            TokenType.ANTI,
304        }
305
306        PLACEHOLDER_PARSERS = {
307            **parser.Parser.PLACEHOLDER_PARSERS,
308            TokenType.PARAMETER: lambda self: (
309                self.expression(exp.Placeholder, this=self._prev.text)
310                if self._match(TokenType.NUMBER) or self._match_set(self.ID_VAR_TOKENS)
311                else None
312            ),
313        }
314
315        def _parse_bracket(
316            self, this: t.Optional[exp.Expression] = None
317        ) -> t.Optional[exp.Expression]:
318            bracket = super()._parse_bracket(this)
319            if isinstance(bracket, exp.Bracket):
320                bracket.set("returns_list_for_maps", True)
321
322            return bracket
323
324        def _parse_map(self) -> exp.ToMap | exp.Map:
325            if self._match(TokenType.L_BRACE, advance=False):
326                return self.expression(exp.ToMap, this=self._parse_bracket())
327
328            args = self._parse_wrapped_csv(self._parse_conjunction)
329            return self.expression(exp.Map, keys=seq_get(args, 0), values=seq_get(args, 1))
330
331        def _parse_types(
332            self, check_func: bool = False, schema: bool = False, allow_identifiers: bool = True
333        ) -> t.Optional[exp.Expression]:
334            this = super()._parse_types(
335                check_func=check_func, schema=schema, allow_identifiers=allow_identifiers
336            )
337
338            # DuckDB treats NUMERIC and DECIMAL without precision as DECIMAL(18, 3)
339            # See: https://duckdb.org/docs/sql/data_types/numeric
340            if (
341                isinstance(this, exp.DataType)
342                and this.is_type("numeric", "decimal")
343                and not this.expressions
344            ):
345                return exp.DataType.build("DECIMAL(18, 3)")
346
347            return this
348
349        def _parse_struct_types(self, type_required: bool = False) -> t.Optional[exp.Expression]:
350            return self._parse_field_def()
351
352        def _pivot_column_names(self, aggregations: t.List[exp.Expression]) -> t.List[str]:
353            if len(aggregations) == 1:
354                return super()._pivot_column_names(aggregations)
355            return pivot_column_names(aggregations, dialect="duckdb")
356
357    class Generator(generator.Generator):
358        PARAMETER_TOKEN = "$"
359        NAMED_PLACEHOLDER_TOKEN = "$"
360        JOIN_HINTS = False
361        TABLE_HINTS = False
362        QUERY_HINTS = False
363        LIMIT_FETCH = "LIMIT"
364        STRUCT_DELIMITER = ("(", ")")
365        RENAME_TABLE_WITH_DB = False
366        NVL2_SUPPORTED = False
367        SEMI_ANTI_JOIN_WITH_SIDE = False
368        TABLESAMPLE_KEYWORDS = "USING SAMPLE"
369        TABLESAMPLE_SEED_KEYWORD = "REPEATABLE"
370        LAST_DAY_SUPPORTS_DATE_PART = False
371        JSON_KEY_VALUE_PAIR_SEP = ","
372        IGNORE_NULLS_IN_FUNC = True
373        JSON_PATH_BRACKETED_KEY_SUPPORTED = False
374        SUPPORTS_CREATE_TABLE_LIKE = False
375        MULTI_ARG_DISTINCT = False
376        CAN_IMPLEMENT_ARRAY_ANY = True
377        SUPPORTS_TO_NUMBER = False
378
379        TRANSFORMS = {
380            **generator.Generator.TRANSFORMS,
381            exp.ApproxDistinct: approx_count_distinct_sql,
382            exp.Array: inline_array_unless_query,
383            exp.ArrayFilter: rename_func("LIST_FILTER"),
384            exp.ArraySize: rename_func("ARRAY_LENGTH"),
385            exp.ArgMax: arg_max_or_min_no_count("ARG_MAX"),
386            exp.ArgMin: arg_max_or_min_no_count("ARG_MIN"),
387            exp.ArraySort: _array_sort_sql,
388            exp.ArraySum: rename_func("LIST_SUM"),
389            exp.BitwiseXor: rename_func("XOR"),
390            exp.CommentColumnConstraint: no_comment_column_constraint_sql,
391            exp.CurrentDate: lambda *_: "CURRENT_DATE",
392            exp.CurrentTime: lambda *_: "CURRENT_TIME",
393            exp.CurrentTimestamp: lambda *_: "CURRENT_TIMESTAMP",
394            exp.DayOfMonth: rename_func("DAYOFMONTH"),
395            exp.DayOfWeek: rename_func("DAYOFWEEK"),
396            exp.DayOfYear: rename_func("DAYOFYEAR"),
397            exp.DataType: _datatype_sql,
398            exp.Date: _date_sql,
399            exp.DateAdd: _date_delta_sql,
400            exp.DateFromParts: rename_func("MAKE_DATE"),
401            exp.DateSub: _date_delta_sql,
402            exp.DateDiff: lambda self, e: self.func(
403                "DATE_DIFF", f"'{e.args.get('unit') or 'DAY'}'", e.expression, e.this
404            ),
405            exp.DateStrToDate: datestrtodate_sql,
406            exp.DateToDi: lambda self,
407            e: f"CAST(STRFTIME({self.sql(e, 'this')}, {DuckDB.DATEINT_FORMAT}) AS INT)",
408            exp.Decode: lambda self, e: encode_decode_sql(self, e, "DECODE", replace=False),
409            exp.DiToDate: lambda self,
410            e: f"CAST(STRPTIME(CAST({self.sql(e, 'this')} AS TEXT), {DuckDB.DATEINT_FORMAT}) AS DATE)",
411            exp.Encode: lambda self, e: encode_decode_sql(self, e, "ENCODE", replace=False),
412            exp.Explode: rename_func("UNNEST"),
413            exp.IntDiv: lambda self, e: self.binary(e, "//"),
414            exp.IsInf: rename_func("ISINF"),
415            exp.IsNan: rename_func("ISNAN"),
416            exp.JSONExtract: arrow_json_extract_sql,
417            exp.JSONExtractScalar: arrow_json_extract_sql,
418            exp.JSONFormat: _json_format_sql,
419            exp.LogicalOr: rename_func("BOOL_OR"),
420            exp.LogicalAnd: rename_func("BOOL_AND"),
421            exp.MonthsBetween: lambda self, e: self.func(
422                "DATEDIFF",
423                "'month'",
424                exp.cast(e.expression, "timestamp", copy=True),
425                exp.cast(e.this, "timestamp", copy=True),
426            ),
427            exp.ParseJSON: rename_func("JSON"),
428            exp.PercentileCont: _rename_unless_within_group("PERCENTILE_CONT", "QUANTILE_CONT"),
429            exp.PercentileDisc: _rename_unless_within_group("PERCENTILE_DISC", "QUANTILE_DISC"),
430            # DuckDB doesn't allow qualified columns inside of PIVOT expressions.
431            # See: https://github.com/duckdb/duckdb/blob/671faf92411182f81dce42ac43de8bfb05d9909e/src/planner/binder/tableref/bind_pivot.cpp#L61-L62
432            exp.Pivot: transforms.preprocess([transforms.unqualify_columns]),
433            exp.RegexpExtract: regexp_extract_sql,
434            exp.RegexpReplace: lambda self, e: self.func(
435                "REGEXP_REPLACE",
436                e.this,
437                e.expression,
438                e.args.get("replacement"),
439                e.args.get("modifiers"),
440            ),
441            exp.RegexpLike: rename_func("REGEXP_MATCHES"),
442            exp.RegexpSplit: rename_func("STR_SPLIT_REGEX"),
443            exp.Rand: rename_func("RANDOM"),
444            exp.SafeDivide: no_safe_divide_sql,
445            exp.Split: rename_func("STR_SPLIT"),
446            exp.SortArray: _sort_array_sql,
447            exp.StrPosition: str_position_sql,
448            exp.StrToDate: lambda self, e: f"CAST({str_to_time_sql(self, e)} AS DATE)",
449            exp.StrToTime: str_to_time_sql,
450            exp.StrToUnix: lambda self, e: self.func(
451                "EPOCH", self.func("STRPTIME", e.this, self.format_time(e))
452            ),
453            exp.Struct: _struct_sql,
454            exp.TimeAdd: _date_delta_sql,
455            exp.Timestamp: no_timestamp_sql,
456            exp.TimestampDiff: lambda self, e: self.func(
457                "DATE_DIFF", exp.Literal.string(e.unit), e.expression, e.this
458            ),
459            exp.TimestampTrunc: timestamptrunc_sql,
460            exp.TimeStrToDate: lambda self, e: self.sql(exp.cast(e.this, "date")),
461            exp.TimeStrToTime: timestrtotime_sql,
462            exp.TimeStrToUnix: lambda self, e: self.func("EPOCH", exp.cast(e.this, "timestamp")),
463            exp.TimeToStr: lambda self, e: self.func("STRFTIME", e.this, self.format_time(e)),
464            exp.TimeToUnix: rename_func("EPOCH"),
465            exp.TsOrDiToDi: lambda self,
466            e: f"CAST(SUBSTR(REPLACE(CAST({self.sql(e, 'this')} AS TEXT), '-', ''), 1, 8) AS INT)",
467            exp.TsOrDsAdd: _ts_or_ds_add_sql,
468            exp.TsOrDsDiff: lambda self, e: self.func(
469                "DATE_DIFF",
470                f"'{e.args.get('unit') or 'DAY'}'",
471                exp.cast(e.expression, "TIMESTAMP"),
472                exp.cast(e.this, "TIMESTAMP"),
473            ),
474            exp.UnixToStr: lambda self, e: self.func(
475                "STRFTIME", self.func("TO_TIMESTAMP", e.this), self.format_time(e)
476            ),
477            exp.UnixToTime: _unix_to_time_sql,
478            exp.UnixToTimeStr: lambda self, e: f"CAST(TO_TIMESTAMP({self.sql(e, 'this')}) AS TEXT)",
479            exp.VariancePop: rename_func("VAR_POP"),
480            exp.WeekOfYear: rename_func("WEEKOFYEAR"),
481            exp.Xor: bool_xor_sql,
482        }
483
484        SUPPORTED_JSON_PATH_PARTS = {
485            exp.JSONPathKey,
486            exp.JSONPathRoot,
487            exp.JSONPathSubscript,
488            exp.JSONPathWildcard,
489        }
490
491        TYPE_MAPPING = {
492            **generator.Generator.TYPE_MAPPING,
493            exp.DataType.Type.BINARY: "BLOB",
494            exp.DataType.Type.CHAR: "TEXT",
495            exp.DataType.Type.FLOAT: "REAL",
496            exp.DataType.Type.NCHAR: "TEXT",
497            exp.DataType.Type.NVARCHAR: "TEXT",
498            exp.DataType.Type.UINT: "UINTEGER",
499            exp.DataType.Type.VARBINARY: "BLOB",
500            exp.DataType.Type.VARCHAR: "TEXT",
501            exp.DataType.Type.TIMESTAMP_S: "TIMESTAMP_S",
502            exp.DataType.Type.TIMESTAMP_MS: "TIMESTAMP_MS",
503            exp.DataType.Type.TIMESTAMP_NS: "TIMESTAMP_NS",
504        }
505
506        STAR_MAPPING = {**generator.Generator.STAR_MAPPING, "except": "EXCLUDE"}
507
508        UNWRAPPED_INTERVAL_VALUES = (exp.Literal, exp.Paren)
509
510        # DuckDB doesn't generally support CREATE TABLE .. properties
511        # https://duckdb.org/docs/sql/statements/create_table.html
512        PROPERTIES_LOCATION = {
513            prop: exp.Properties.Location.UNSUPPORTED
514            for prop in generator.Generator.PROPERTIES_LOCATION
515        }
516
517        # There are a few exceptions (e.g. temporary tables) which are supported or
518        # can be transpiled to DuckDB, so we explicitly override them accordingly
519        PROPERTIES_LOCATION[exp.LikeProperty] = exp.Properties.Location.POST_SCHEMA
520        PROPERTIES_LOCATION[exp.TemporaryProperty] = exp.Properties.Location.POST_CREATE
521
522        def timefromparts_sql(self, expression: exp.TimeFromParts) -> str:
523            nano = expression.args.get("nano")
524            if nano is not None:
525                expression.set(
526                    "sec", expression.args["sec"] + nano.pop() / exp.Literal.number(1000000000.0)
527                )
528
529            return rename_func("MAKE_TIME")(self, expression)
530
531        def timestampfromparts_sql(self, expression: exp.TimestampFromParts) -> str:
532            sec = expression.args["sec"]
533
534            milli = expression.args.get("milli")
535            if milli is not None:
536                sec += milli.pop() / exp.Literal.number(1000.0)
537
538            nano = expression.args.get("nano")
539            if nano is not None:
540                sec += nano.pop() / exp.Literal.number(1000000000.0)
541
542            if milli or nano:
543                expression.set("sec", sec)
544
545            return rename_func("MAKE_TIMESTAMP")(self, expression)
546
547        def tablesample_sql(
548            self,
549            expression: exp.TableSample,
550            sep: str = " AS ",
551            tablesample_keyword: t.Optional[str] = None,
552        ) -> str:
553            if not isinstance(expression.parent, exp.Select):
554                # This sample clause only applies to a single source, not the entire resulting relation
555                tablesample_keyword = "TABLESAMPLE"
556
557            return super().tablesample_sql(
558                expression, sep=sep, tablesample_keyword=tablesample_keyword
559            )
560
561        def interval_sql(self, expression: exp.Interval) -> str:
562            multiplier: t.Optional[int] = None
563            unit = expression.text("unit").lower()
564
565            if unit.startswith("week"):
566                multiplier = 7
567            if unit.startswith("quarter"):
568                multiplier = 90
569
570            if multiplier:
571                return f"({multiplier} * {super().interval_sql(exp.Interval(this=expression.this, unit=exp.var('DAY')))})"
572
573            return super().interval_sql(expression)
574
575        def columndef_sql(self, expression: exp.ColumnDef, sep: str = " ") -> str:
576            if isinstance(expression.parent, exp.UserDefinedFunction):
577                return self.sql(expression, "this")
578            return super().columndef_sql(expression, sep)
579
580        def join_sql(self, expression: exp.Join) -> str:
581            if (
582                expression.side == "LEFT"
583                and not expression.args.get("on")
584                and isinstance(expression.this, exp.Unnest)
585            ):
586                # Some dialects support `LEFT JOIN UNNEST(...)` without an explicit ON clause
587                # DuckDB doesn't, but we can just add a dummy ON clause that is always true
588                return super().join_sql(expression.on(exp.true()))
589
590            return super().join_sql(expression)
591
592        def generateseries_sql(self, expression: exp.GenerateSeries) -> str:
593            # GENERATE_SERIES(a, b) -> [a, b], RANGE(a, b) -> [a, b)
594            if expression.args.get("is_end_exclusive"):
595                expression.set("is_end_exclusive", None)
596                return rename_func("RANGE")(self, expression)
597
598            return super().generateseries_sql(expression)
599
600        def bracket_sql(self, expression: exp.Bracket) -> str:
601            this = expression.this
602            if isinstance(this, exp.Array):
603                this.replace(exp.paren(this))
604
605            bracket = super().bracket_sql(expression)
606
607            if not expression.args.get("returns_list_for_maps"):
608                if not this.type:
609                    from sqlglot.optimizer.annotate_types import annotate_types
610
611                    this = annotate_types(this)
612
613                if this.is_type(exp.DataType.Type.MAP):
614                    bracket = f"({bracket})[1]"
615
616            return bracket
class DuckDB(sqlglot.dialects.dialect.Dialect):
169class DuckDB(Dialect):
170    NULL_ORDERING = "nulls_are_last"
171    SUPPORTS_USER_DEFINED_TYPES = False
172    SAFE_DIVISION = True
173    INDEX_OFFSET = 1
174    CONCAT_COALESCE = True
175
176    # https://duckdb.org/docs/sql/introduction.html#creating-a-new-table
177    NORMALIZATION_STRATEGY = NormalizationStrategy.CASE_INSENSITIVE
178
179    def to_json_path(self, path: t.Optional[exp.Expression]) -> t.Optional[exp.Expression]:
180        if isinstance(path, exp.Literal):
181            # DuckDB also supports the JSON pointer syntax, where every path starts with a `/`.
182            # Additionally, it allows accessing the back of lists using the `[#-i]` syntax.
183            # This check ensures we'll avoid trying to parse these as JSON paths, which can
184            # either result in a noisy warning or in an invalid representation of the path.
185            path_text = path.name
186            if path_text.startswith("/") or "[#" in path_text:
187                return path
188
189        return super().to_json_path(path)
190
191    class Tokenizer(tokens.Tokenizer):
192        HEREDOC_STRINGS = ["$"]
193
194        HEREDOC_TAG_IS_IDENTIFIER = True
195        HEREDOC_STRING_ALTERNATIVE = TokenType.PARAMETER
196
197        KEYWORDS = {
198            **tokens.Tokenizer.KEYWORDS,
199            "//": TokenType.DIV,
200            "ATTACH": TokenType.COMMAND,
201            "BINARY": TokenType.VARBINARY,
202            "BITSTRING": TokenType.BIT,
203            "BPCHAR": TokenType.TEXT,
204            "CHAR": TokenType.TEXT,
205            "CHARACTER VARYING": TokenType.TEXT,
206            "EXCLUDE": TokenType.EXCEPT,
207            "LOGICAL": TokenType.BOOLEAN,
208            "ONLY": TokenType.ONLY,
209            "PIVOT_WIDER": TokenType.PIVOT,
210            "POSITIONAL": TokenType.POSITIONAL,
211            "SIGNED": TokenType.INT,
212            "STRING": TokenType.VARCHAR,
213            "UBIGINT": TokenType.UBIGINT,
214            "UINTEGER": TokenType.UINT,
215            "USMALLINT": TokenType.USMALLINT,
216            "UTINYINT": TokenType.UTINYINT,
217            "TIMESTAMP_S": TokenType.TIMESTAMP_S,
218            "TIMESTAMP_MS": TokenType.TIMESTAMP_MS,
219            "TIMESTAMP_NS": TokenType.TIMESTAMP_NS,
220            "TIMESTAMP_US": TokenType.TIMESTAMP,
221        }
222
223        SINGLE_TOKENS = {
224            **tokens.Tokenizer.SINGLE_TOKENS,
225            "$": TokenType.PARAMETER,
226        }
227
228    class Parser(parser.Parser):
229        BITWISE = {
230            **parser.Parser.BITWISE,
231            TokenType.TILDA: exp.RegexpLike,
232        }
233
234        FUNCTIONS_WITH_ALIASED_ARGS = {*parser.Parser.FUNCTIONS_WITH_ALIASED_ARGS, "STRUCT_PACK"}
235
236        FUNCTIONS = {
237            **parser.Parser.FUNCTIONS,
238            "ARRAY_HAS": exp.ArrayContains.from_arg_list,
239            "ARRAY_REVERSE_SORT": _build_sort_array_desc,
240            "ARRAY_SORT": exp.SortArray.from_arg_list,
241            "DATEDIFF": _build_date_diff,
242            "DATE_DIFF": _build_date_diff,
243            "DATE_TRUNC": date_trunc_to_time,
244            "DATETRUNC": date_trunc_to_time,
245            "DECODE": lambda args: exp.Decode(
246                this=seq_get(args, 0), charset=exp.Literal.string("utf-8")
247            ),
248            "ENCODE": lambda args: exp.Encode(
249                this=seq_get(args, 0), charset=exp.Literal.string("utf-8")
250            ),
251            "EPOCH": exp.TimeToUnix.from_arg_list,
252            "EPOCH_MS": lambda args: exp.UnixToTime(
253                this=seq_get(args, 0), scale=exp.UnixToTime.MILLIS
254            ),
255            "JSON": exp.ParseJSON.from_arg_list,
256            "JSON_EXTRACT_PATH": parser.build_extract_json_with_path(exp.JSONExtract),
257            "JSON_EXTRACT_STRING": parser.build_extract_json_with_path(exp.JSONExtractScalar),
258            "LIST_HAS": exp.ArrayContains.from_arg_list,
259            "LIST_REVERSE_SORT": _build_sort_array_desc,
260            "LIST_SORT": exp.SortArray.from_arg_list,
261            "LIST_VALUE": exp.Array.from_arg_list,
262            "MAKE_TIME": exp.TimeFromParts.from_arg_list,
263            "MAKE_TIMESTAMP": _build_make_timestamp,
264            "MEDIAN": lambda args: exp.PercentileCont(
265                this=seq_get(args, 0), expression=exp.Literal.number(0.5)
266            ),
267            "QUANTILE_CONT": exp.PercentileCont.from_arg_list,
268            "QUANTILE_DISC": exp.PercentileDisc.from_arg_list,
269            "REGEXP_EXTRACT": lambda args: exp.RegexpExtract(
270                this=seq_get(args, 0), expression=seq_get(args, 1), group=seq_get(args, 2)
271            ),
272            "REGEXP_MATCHES": exp.RegexpLike.from_arg_list,
273            "REGEXP_REPLACE": lambda args: exp.RegexpReplace(
274                this=seq_get(args, 0),
275                expression=seq_get(args, 1),
276                replacement=seq_get(args, 2),
277                modifiers=seq_get(args, 3),
278            ),
279            "STRFTIME": build_formatted_time(exp.TimeToStr, "duckdb"),
280            "STRING_SPLIT": exp.Split.from_arg_list,
281            "STRING_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
282            "STRING_TO_ARRAY": exp.Split.from_arg_list,
283            "STRPTIME": build_formatted_time(exp.StrToTime, "duckdb"),
284            "STRUCT_PACK": exp.Struct.from_arg_list,
285            "STR_SPLIT": exp.Split.from_arg_list,
286            "STR_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
287            "TO_TIMESTAMP": exp.UnixToTime.from_arg_list,
288            "UNNEST": exp.Explode.from_arg_list,
289            "XOR": binary_from_function(exp.BitwiseXor),
290            "GENERATE_SERIES": _build_generate_series(),
291            "RANGE": _build_generate_series(end_exclusive=True),
292        }
293
294        FUNCTION_PARSERS = parser.Parser.FUNCTION_PARSERS.copy()
295        FUNCTION_PARSERS.pop("DECODE")
296
297        NO_PAREN_FUNCTION_PARSERS = {
298            **parser.Parser.NO_PAREN_FUNCTION_PARSERS,
299            "MAP": lambda self: self._parse_map(),
300        }
301
302        TABLE_ALIAS_TOKENS = parser.Parser.TABLE_ALIAS_TOKENS - {
303            TokenType.SEMI,
304            TokenType.ANTI,
305        }
306
307        PLACEHOLDER_PARSERS = {
308            **parser.Parser.PLACEHOLDER_PARSERS,
309            TokenType.PARAMETER: lambda self: (
310                self.expression(exp.Placeholder, this=self._prev.text)
311                if self._match(TokenType.NUMBER) or self._match_set(self.ID_VAR_TOKENS)
312                else None
313            ),
314        }
315
316        def _parse_bracket(
317            self, this: t.Optional[exp.Expression] = None
318        ) -> t.Optional[exp.Expression]:
319            bracket = super()._parse_bracket(this)
320            if isinstance(bracket, exp.Bracket):
321                bracket.set("returns_list_for_maps", True)
322
323            return bracket
324
325        def _parse_map(self) -> exp.ToMap | exp.Map:
326            if self._match(TokenType.L_BRACE, advance=False):
327                return self.expression(exp.ToMap, this=self._parse_bracket())
328
329            args = self._parse_wrapped_csv(self._parse_conjunction)
330            return self.expression(exp.Map, keys=seq_get(args, 0), values=seq_get(args, 1))
331
332        def _parse_types(
333            self, check_func: bool = False, schema: bool = False, allow_identifiers: bool = True
334        ) -> t.Optional[exp.Expression]:
335            this = super()._parse_types(
336                check_func=check_func, schema=schema, allow_identifiers=allow_identifiers
337            )
338
339            # DuckDB treats NUMERIC and DECIMAL without precision as DECIMAL(18, 3)
340            # See: https://duckdb.org/docs/sql/data_types/numeric
341            if (
342                isinstance(this, exp.DataType)
343                and this.is_type("numeric", "decimal")
344                and not this.expressions
345            ):
346                return exp.DataType.build("DECIMAL(18, 3)")
347
348            return this
349
350        def _parse_struct_types(self, type_required: bool = False) -> t.Optional[exp.Expression]:
351            return self._parse_field_def()
352
353        def _pivot_column_names(self, aggregations: t.List[exp.Expression]) -> t.List[str]:
354            if len(aggregations) == 1:
355                return super()._pivot_column_names(aggregations)
356            return pivot_column_names(aggregations, dialect="duckdb")
357
358    class Generator(generator.Generator):
359        PARAMETER_TOKEN = "$"
360        NAMED_PLACEHOLDER_TOKEN = "$"
361        JOIN_HINTS = False
362        TABLE_HINTS = False
363        QUERY_HINTS = False
364        LIMIT_FETCH = "LIMIT"
365        STRUCT_DELIMITER = ("(", ")")
366        RENAME_TABLE_WITH_DB = False
367        NVL2_SUPPORTED = False
368        SEMI_ANTI_JOIN_WITH_SIDE = False
369        TABLESAMPLE_KEYWORDS = "USING SAMPLE"
370        TABLESAMPLE_SEED_KEYWORD = "REPEATABLE"
371        LAST_DAY_SUPPORTS_DATE_PART = False
372        JSON_KEY_VALUE_PAIR_SEP = ","
373        IGNORE_NULLS_IN_FUNC = True
374        JSON_PATH_BRACKETED_KEY_SUPPORTED = False
375        SUPPORTS_CREATE_TABLE_LIKE = False
376        MULTI_ARG_DISTINCT = False
377        CAN_IMPLEMENT_ARRAY_ANY = True
378        SUPPORTS_TO_NUMBER = False
379
380        TRANSFORMS = {
381            **generator.Generator.TRANSFORMS,
382            exp.ApproxDistinct: approx_count_distinct_sql,
383            exp.Array: inline_array_unless_query,
384            exp.ArrayFilter: rename_func("LIST_FILTER"),
385            exp.ArraySize: rename_func("ARRAY_LENGTH"),
386            exp.ArgMax: arg_max_or_min_no_count("ARG_MAX"),
387            exp.ArgMin: arg_max_or_min_no_count("ARG_MIN"),
388            exp.ArraySort: _array_sort_sql,
389            exp.ArraySum: rename_func("LIST_SUM"),
390            exp.BitwiseXor: rename_func("XOR"),
391            exp.CommentColumnConstraint: no_comment_column_constraint_sql,
392            exp.CurrentDate: lambda *_: "CURRENT_DATE",
393            exp.CurrentTime: lambda *_: "CURRENT_TIME",
394            exp.CurrentTimestamp: lambda *_: "CURRENT_TIMESTAMP",
395            exp.DayOfMonth: rename_func("DAYOFMONTH"),
396            exp.DayOfWeek: rename_func("DAYOFWEEK"),
397            exp.DayOfYear: rename_func("DAYOFYEAR"),
398            exp.DataType: _datatype_sql,
399            exp.Date: _date_sql,
400            exp.DateAdd: _date_delta_sql,
401            exp.DateFromParts: rename_func("MAKE_DATE"),
402            exp.DateSub: _date_delta_sql,
403            exp.DateDiff: lambda self, e: self.func(
404                "DATE_DIFF", f"'{e.args.get('unit') or 'DAY'}'", e.expression, e.this
405            ),
406            exp.DateStrToDate: datestrtodate_sql,
407            exp.DateToDi: lambda self,
408            e: f"CAST(STRFTIME({self.sql(e, 'this')}, {DuckDB.DATEINT_FORMAT}) AS INT)",
409            exp.Decode: lambda self, e: encode_decode_sql(self, e, "DECODE", replace=False),
410            exp.DiToDate: lambda self,
411            e: f"CAST(STRPTIME(CAST({self.sql(e, 'this')} AS TEXT), {DuckDB.DATEINT_FORMAT}) AS DATE)",
412            exp.Encode: lambda self, e: encode_decode_sql(self, e, "ENCODE", replace=False),
413            exp.Explode: rename_func("UNNEST"),
414            exp.IntDiv: lambda self, e: self.binary(e, "//"),
415            exp.IsInf: rename_func("ISINF"),
416            exp.IsNan: rename_func("ISNAN"),
417            exp.JSONExtract: arrow_json_extract_sql,
418            exp.JSONExtractScalar: arrow_json_extract_sql,
419            exp.JSONFormat: _json_format_sql,
420            exp.LogicalOr: rename_func("BOOL_OR"),
421            exp.LogicalAnd: rename_func("BOOL_AND"),
422            exp.MonthsBetween: lambda self, e: self.func(
423                "DATEDIFF",
424                "'month'",
425                exp.cast(e.expression, "timestamp", copy=True),
426                exp.cast(e.this, "timestamp", copy=True),
427            ),
428            exp.ParseJSON: rename_func("JSON"),
429            exp.PercentileCont: _rename_unless_within_group("PERCENTILE_CONT", "QUANTILE_CONT"),
430            exp.PercentileDisc: _rename_unless_within_group("PERCENTILE_DISC", "QUANTILE_DISC"),
431            # DuckDB doesn't allow qualified columns inside of PIVOT expressions.
432            # See: https://github.com/duckdb/duckdb/blob/671faf92411182f81dce42ac43de8bfb05d9909e/src/planner/binder/tableref/bind_pivot.cpp#L61-L62
433            exp.Pivot: transforms.preprocess([transforms.unqualify_columns]),
434            exp.RegexpExtract: regexp_extract_sql,
435            exp.RegexpReplace: lambda self, e: self.func(
436                "REGEXP_REPLACE",
437                e.this,
438                e.expression,
439                e.args.get("replacement"),
440                e.args.get("modifiers"),
441            ),
442            exp.RegexpLike: rename_func("REGEXP_MATCHES"),
443            exp.RegexpSplit: rename_func("STR_SPLIT_REGEX"),
444            exp.Rand: rename_func("RANDOM"),
445            exp.SafeDivide: no_safe_divide_sql,
446            exp.Split: rename_func("STR_SPLIT"),
447            exp.SortArray: _sort_array_sql,
448            exp.StrPosition: str_position_sql,
449            exp.StrToDate: lambda self, e: f"CAST({str_to_time_sql(self, e)} AS DATE)",
450            exp.StrToTime: str_to_time_sql,
451            exp.StrToUnix: lambda self, e: self.func(
452                "EPOCH", self.func("STRPTIME", e.this, self.format_time(e))
453            ),
454            exp.Struct: _struct_sql,
455            exp.TimeAdd: _date_delta_sql,
456            exp.Timestamp: no_timestamp_sql,
457            exp.TimestampDiff: lambda self, e: self.func(
458                "DATE_DIFF", exp.Literal.string(e.unit), e.expression, e.this
459            ),
460            exp.TimestampTrunc: timestamptrunc_sql,
461            exp.TimeStrToDate: lambda self, e: self.sql(exp.cast(e.this, "date")),
462            exp.TimeStrToTime: timestrtotime_sql,
463            exp.TimeStrToUnix: lambda self, e: self.func("EPOCH", exp.cast(e.this, "timestamp")),
464            exp.TimeToStr: lambda self, e: self.func("STRFTIME", e.this, self.format_time(e)),
465            exp.TimeToUnix: rename_func("EPOCH"),
466            exp.TsOrDiToDi: lambda self,
467            e: f"CAST(SUBSTR(REPLACE(CAST({self.sql(e, 'this')} AS TEXT), '-', ''), 1, 8) AS INT)",
468            exp.TsOrDsAdd: _ts_or_ds_add_sql,
469            exp.TsOrDsDiff: lambda self, e: self.func(
470                "DATE_DIFF",
471                f"'{e.args.get('unit') or 'DAY'}'",
472                exp.cast(e.expression, "TIMESTAMP"),
473                exp.cast(e.this, "TIMESTAMP"),
474            ),
475            exp.UnixToStr: lambda self, e: self.func(
476                "STRFTIME", self.func("TO_TIMESTAMP", e.this), self.format_time(e)
477            ),
478            exp.UnixToTime: _unix_to_time_sql,
479            exp.UnixToTimeStr: lambda self, e: f"CAST(TO_TIMESTAMP({self.sql(e, 'this')}) AS TEXT)",
480            exp.VariancePop: rename_func("VAR_POP"),
481            exp.WeekOfYear: rename_func("WEEKOFYEAR"),
482            exp.Xor: bool_xor_sql,
483        }
484
485        SUPPORTED_JSON_PATH_PARTS = {
486            exp.JSONPathKey,
487            exp.JSONPathRoot,
488            exp.JSONPathSubscript,
489            exp.JSONPathWildcard,
490        }
491
492        TYPE_MAPPING = {
493            **generator.Generator.TYPE_MAPPING,
494            exp.DataType.Type.BINARY: "BLOB",
495            exp.DataType.Type.CHAR: "TEXT",
496            exp.DataType.Type.FLOAT: "REAL",
497            exp.DataType.Type.NCHAR: "TEXT",
498            exp.DataType.Type.NVARCHAR: "TEXT",
499            exp.DataType.Type.UINT: "UINTEGER",
500            exp.DataType.Type.VARBINARY: "BLOB",
501            exp.DataType.Type.VARCHAR: "TEXT",
502            exp.DataType.Type.TIMESTAMP_S: "TIMESTAMP_S",
503            exp.DataType.Type.TIMESTAMP_MS: "TIMESTAMP_MS",
504            exp.DataType.Type.TIMESTAMP_NS: "TIMESTAMP_NS",
505        }
506
507        STAR_MAPPING = {**generator.Generator.STAR_MAPPING, "except": "EXCLUDE"}
508
509        UNWRAPPED_INTERVAL_VALUES = (exp.Literal, exp.Paren)
510
511        # DuckDB doesn't generally support CREATE TABLE .. properties
512        # https://duckdb.org/docs/sql/statements/create_table.html
513        PROPERTIES_LOCATION = {
514            prop: exp.Properties.Location.UNSUPPORTED
515            for prop in generator.Generator.PROPERTIES_LOCATION
516        }
517
518        # There are a few exceptions (e.g. temporary tables) which are supported or
519        # can be transpiled to DuckDB, so we explicitly override them accordingly
520        PROPERTIES_LOCATION[exp.LikeProperty] = exp.Properties.Location.POST_SCHEMA
521        PROPERTIES_LOCATION[exp.TemporaryProperty] = exp.Properties.Location.POST_CREATE
522
523        def timefromparts_sql(self, expression: exp.TimeFromParts) -> str:
524            nano = expression.args.get("nano")
525            if nano is not None:
526                expression.set(
527                    "sec", expression.args["sec"] + nano.pop() / exp.Literal.number(1000000000.0)
528                )
529
530            return rename_func("MAKE_TIME")(self, expression)
531
532        def timestampfromparts_sql(self, expression: exp.TimestampFromParts) -> str:
533            sec = expression.args["sec"]
534
535            milli = expression.args.get("milli")
536            if milli is not None:
537                sec += milli.pop() / exp.Literal.number(1000.0)
538
539            nano = expression.args.get("nano")
540            if nano is not None:
541                sec += nano.pop() / exp.Literal.number(1000000000.0)
542
543            if milli or nano:
544                expression.set("sec", sec)
545
546            return rename_func("MAKE_TIMESTAMP")(self, expression)
547
548        def tablesample_sql(
549            self,
550            expression: exp.TableSample,
551            sep: str = " AS ",
552            tablesample_keyword: t.Optional[str] = None,
553        ) -> str:
554            if not isinstance(expression.parent, exp.Select):
555                # This sample clause only applies to a single source, not the entire resulting relation
556                tablesample_keyword = "TABLESAMPLE"
557
558            return super().tablesample_sql(
559                expression, sep=sep, tablesample_keyword=tablesample_keyword
560            )
561
562        def interval_sql(self, expression: exp.Interval) -> str:
563            multiplier: t.Optional[int] = None
564            unit = expression.text("unit").lower()
565
566            if unit.startswith("week"):
567                multiplier = 7
568            if unit.startswith("quarter"):
569                multiplier = 90
570
571            if multiplier:
572                return f"({multiplier} * {super().interval_sql(exp.Interval(this=expression.this, unit=exp.var('DAY')))})"
573
574            return super().interval_sql(expression)
575
576        def columndef_sql(self, expression: exp.ColumnDef, sep: str = " ") -> str:
577            if isinstance(expression.parent, exp.UserDefinedFunction):
578                return self.sql(expression, "this")
579            return super().columndef_sql(expression, sep)
580
581        def join_sql(self, expression: exp.Join) -> str:
582            if (
583                expression.side == "LEFT"
584                and not expression.args.get("on")
585                and isinstance(expression.this, exp.Unnest)
586            ):
587                # Some dialects support `LEFT JOIN UNNEST(...)` without an explicit ON clause
588                # DuckDB doesn't, but we can just add a dummy ON clause that is always true
589                return super().join_sql(expression.on(exp.true()))
590
591            return super().join_sql(expression)
592
593        def generateseries_sql(self, expression: exp.GenerateSeries) -> str:
594            # GENERATE_SERIES(a, b) -> [a, b], RANGE(a, b) -> [a, b)
595            if expression.args.get("is_end_exclusive"):
596                expression.set("is_end_exclusive", None)
597                return rename_func("RANGE")(self, expression)
598
599            return super().generateseries_sql(expression)
600
601        def bracket_sql(self, expression: exp.Bracket) -> str:
602            this = expression.this
603            if isinstance(this, exp.Array):
604                this.replace(exp.paren(this))
605
606            bracket = super().bracket_sql(expression)
607
608            if not expression.args.get("returns_list_for_maps"):
609                if not this.type:
610                    from sqlglot.optimizer.annotate_types import annotate_types
611
612                    this = annotate_types(this)
613
614                if this.is_type(exp.DataType.Type.MAP):
615                    bracket = f"({bracket})[1]"
616
617            return bracket
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.

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]:
179    def to_json_path(self, path: t.Optional[exp.Expression]) -> t.Optional[exp.Expression]:
180        if isinstance(path, exp.Literal):
181            # DuckDB also supports the JSON pointer syntax, where every path starts with a `/`.
182            # Additionally, it allows accessing the back of lists using the `[#-i]` syntax.
183            # This check ensures we'll avoid trying to parse these as JSON paths, which can
184            # either result in a noisy warning or in an invalid representation of the path.
185            path_text = path.name
186            if path_text.startswith("/") or "[#" in path_text:
187                return path
188
189        return super().to_json_path(path)
tokenizer_class = <class 'DuckDB.Tokenizer'>
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 = {}
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):
191    class Tokenizer(tokens.Tokenizer):
192        HEREDOC_STRINGS = ["$"]
193
194        HEREDOC_TAG_IS_IDENTIFIER = True
195        HEREDOC_STRING_ALTERNATIVE = TokenType.PARAMETER
196
197        KEYWORDS = {
198            **tokens.Tokenizer.KEYWORDS,
199            "//": TokenType.DIV,
200            "ATTACH": TokenType.COMMAND,
201            "BINARY": TokenType.VARBINARY,
202            "BITSTRING": TokenType.BIT,
203            "BPCHAR": TokenType.TEXT,
204            "CHAR": TokenType.TEXT,
205            "CHARACTER VARYING": TokenType.TEXT,
206            "EXCLUDE": TokenType.EXCEPT,
207            "LOGICAL": TokenType.BOOLEAN,
208            "ONLY": TokenType.ONLY,
209            "PIVOT_WIDER": TokenType.PIVOT,
210            "POSITIONAL": TokenType.POSITIONAL,
211            "SIGNED": TokenType.INT,
212            "STRING": TokenType.VARCHAR,
213            "UBIGINT": TokenType.UBIGINT,
214            "UINTEGER": TokenType.UINT,
215            "USMALLINT": TokenType.USMALLINT,
216            "UTINYINT": TokenType.UTINYINT,
217            "TIMESTAMP_S": TokenType.TIMESTAMP_S,
218            "TIMESTAMP_MS": TokenType.TIMESTAMP_MS,
219            "TIMESTAMP_NS": TokenType.TIMESTAMP_NS,
220            "TIMESTAMP_US": TokenType.TIMESTAMP,
221        }
222
223        SINGLE_TOKENS = {
224            **tokens.Tokenizer.SINGLE_TOKENS,
225            "$": TokenType.PARAMETER,
226        }
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.HINT: 'HINT'>, '==': <TokenType.EQ: 'EQ'>, '::': <TokenType.DCOLON: 'DCOLON'>, '||': <TokenType.DPIPE: 'DPIPE'>, '>=': <TokenType.GTE: 'GTE'>, '<=': <TokenType.LTE: 'LTE'>, '<>': <TokenType.NEQ: 'NEQ'>, '!=': <TokenType.NEQ: 'NEQ'>, ':=': <TokenType.COLON_EQ: 'COLON_EQ'>, '<=>': <TokenType.NULLSAFE_EQ: 'NULLSAFE_EQ'>, '->': <TokenType.ARROW: 'ARROW'>, '->>': <TokenType.DARROW: 'DARROW'>, '=>': <TokenType.FARROW: 'FARROW'>, '#>': <TokenType.HASH_ARROW: 'HASH_ARROW'>, '#>>': <TokenType.DHASH_ARROW: 'DHASH_ARROW'>, '<->': <TokenType.LR_ARROW: 'LR_ARROW'>, '&&': <TokenType.DAMP: 'DAMP'>, '??': <TokenType.DQMARK: 'DQMARK'>, 'ALL': <TokenType.ALL: 'ALL'>, 'ALWAYS': <TokenType.ALWAYS: 'ALWAYS'>, 'AND': <TokenType.AND: 'AND'>, 'ANTI': <TokenType.ANTI: 'ANTI'>, 'ANY': <TokenType.ANY: 'ANY'>, 'ASC': <TokenType.ASC: 'ASC'>, 'AS': <TokenType.ALIAS: 'ALIAS'>, 'ASOF': <TokenType.ASOF: 'ASOF'>, 'AUTOINCREMENT': <TokenType.AUTO_INCREMENT: 'AUTO_INCREMENT'>, 'AUTO_INCREMENT': <TokenType.AUTO_INCREMENT: 'AUTO_INCREMENT'>, 'BEGIN': <TokenType.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'>, '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'>, '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'>, 'MAP': <TokenType.MAP: 'MAP'>, 'NULLABLE': <TokenType.NULLABLE: 'NULLABLE'>, 'NUMBER': <TokenType.DECIMAL: 'DECIMAL'>, 'NUMERIC': <TokenType.DECIMAL: 'DECIMAL'>, 'FIXED': <TokenType.DECIMAL: 'DECIMAL'>, 'REAL': <TokenType.FLOAT: 'FLOAT'>, 'FLOAT': <TokenType.FLOAT: 'FLOAT'>, 'FLOAT4': <TokenType.FLOAT: 'FLOAT'>, 'FLOAT8': <TokenType.DOUBLE: 'DOUBLE'>, 'DOUBLE': <TokenType.DOUBLE: 'DOUBLE'>, 'DOUBLE PRECISION': <TokenType.DOUBLE: 'DOUBLE'>, 'JSON': <TokenType.JSON: 'JSON'>, 'CHAR': <TokenType.TEXT: 'TEXT'>, 'CHARACTER': <TokenType.CHAR: 'CHAR'>, 'NCHAR': <TokenType.NCHAR: 'NCHAR'>, 'VARCHAR': <TokenType.VARCHAR: 'VARCHAR'>, 'VARCHAR2': <TokenType.VARCHAR: 'VARCHAR'>, 'NVARCHAR': <TokenType.NVARCHAR: 'NVARCHAR'>, 'NVARCHAR2': <TokenType.NVARCHAR: 'NVARCHAR'>, 'BPCHAR': <TokenType.TEXT: 'TEXT'>, 'STR': <TokenType.TEXT: 'TEXT'>, 'STRING': <TokenType.VARCHAR: 'VARCHAR'>, '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'>, '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'>, 'STRUCT': <TokenType.STRUCT: 'STRUCT'>, 'SEQUENCE': <TokenType.SEQUENCE: 'SEQUENCE'>, 'VARIANT': <TokenType.VARIANT: 'VARIANT'>, 'ALTER': <TokenType.ALTER: 'ALTER'>, 'ANALYZE': <TokenType.COMMAND: 'COMMAND'>, 'CALL': <TokenType.COMMAND: 'COMMAND'>, 'COMMENT': <TokenType.COMMENT: 'COMMENT'>, 'COPY': <TokenType.COMMAND: 'COMMAND'>, 'EXPLAIN': <TokenType.COMMAND: 'COMMAND'>, 'GRANT': <TokenType.COMMAND: 'COMMAND'>, 'OPTIMIZE': <TokenType.COMMAND: 'COMMAND'>, 'PREPARE': <TokenType.COMMAND: 'COMMAND'>, 'VACUUM': <TokenType.COMMAND: 'COMMAND'>, 'USER-DEFINED': <TokenType.USERDEFINED: 'USERDEFINED'>, 'FOR VERSION': <TokenType.VERSION_SNAPSHOT: 'VERSION_SNAPSHOT'>, 'FOR TIMESTAMP': <TokenType.TIMESTAMP_SNAPSHOT: 'TIMESTAMP_SNAPSHOT'>, '//': <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'>, 'UBIGINT': <TokenType.UBIGINT: 'UBIGINT'>, 'UINTEGER': <TokenType.UINT: 'UINT'>, 'USMALLINT': <TokenType.USMALLINT: 'USMALLINT'>, 'UTINYINT': <TokenType.UTINYINT: 'UTINYINT'>, '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'>}
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.QUOTE: 'QUOTE'>, '`': <TokenType.IDENTIFIER: 'IDENTIFIER'>, '"': <TokenType.IDENTIFIER: 'IDENTIFIER'>, '#': <TokenType.HASH: 'HASH'>, '$': <TokenType.PARAMETER: 'PARAMETER'>}
class DuckDB.Parser(sqlglot.parser.Parser):
228    class Parser(parser.Parser):
229        BITWISE = {
230            **parser.Parser.BITWISE,
231            TokenType.TILDA: exp.RegexpLike,
232        }
233
234        FUNCTIONS_WITH_ALIASED_ARGS = {*parser.Parser.FUNCTIONS_WITH_ALIASED_ARGS, "STRUCT_PACK"}
235
236        FUNCTIONS = {
237            **parser.Parser.FUNCTIONS,
238            "ARRAY_HAS": exp.ArrayContains.from_arg_list,
239            "ARRAY_REVERSE_SORT": _build_sort_array_desc,
240            "ARRAY_SORT": exp.SortArray.from_arg_list,
241            "DATEDIFF": _build_date_diff,
242            "DATE_DIFF": _build_date_diff,
243            "DATE_TRUNC": date_trunc_to_time,
244            "DATETRUNC": date_trunc_to_time,
245            "DECODE": lambda args: exp.Decode(
246                this=seq_get(args, 0), charset=exp.Literal.string("utf-8")
247            ),
248            "ENCODE": lambda args: exp.Encode(
249                this=seq_get(args, 0), charset=exp.Literal.string("utf-8")
250            ),
251            "EPOCH": exp.TimeToUnix.from_arg_list,
252            "EPOCH_MS": lambda args: exp.UnixToTime(
253                this=seq_get(args, 0), scale=exp.UnixToTime.MILLIS
254            ),
255            "JSON": exp.ParseJSON.from_arg_list,
256            "JSON_EXTRACT_PATH": parser.build_extract_json_with_path(exp.JSONExtract),
257            "JSON_EXTRACT_STRING": parser.build_extract_json_with_path(exp.JSONExtractScalar),
258            "LIST_HAS": exp.ArrayContains.from_arg_list,
259            "LIST_REVERSE_SORT": _build_sort_array_desc,
260            "LIST_SORT": exp.SortArray.from_arg_list,
261            "LIST_VALUE": exp.Array.from_arg_list,
262            "MAKE_TIME": exp.TimeFromParts.from_arg_list,
263            "MAKE_TIMESTAMP": _build_make_timestamp,
264            "MEDIAN": lambda args: exp.PercentileCont(
265                this=seq_get(args, 0), expression=exp.Literal.number(0.5)
266            ),
267            "QUANTILE_CONT": exp.PercentileCont.from_arg_list,
268            "QUANTILE_DISC": exp.PercentileDisc.from_arg_list,
269            "REGEXP_EXTRACT": lambda args: exp.RegexpExtract(
270                this=seq_get(args, 0), expression=seq_get(args, 1), group=seq_get(args, 2)
271            ),
272            "REGEXP_MATCHES": exp.RegexpLike.from_arg_list,
273            "REGEXP_REPLACE": lambda args: exp.RegexpReplace(
274                this=seq_get(args, 0),
275                expression=seq_get(args, 1),
276                replacement=seq_get(args, 2),
277                modifiers=seq_get(args, 3),
278            ),
279            "STRFTIME": build_formatted_time(exp.TimeToStr, "duckdb"),
280            "STRING_SPLIT": exp.Split.from_arg_list,
281            "STRING_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
282            "STRING_TO_ARRAY": exp.Split.from_arg_list,
283            "STRPTIME": build_formatted_time(exp.StrToTime, "duckdb"),
284            "STRUCT_PACK": exp.Struct.from_arg_list,
285            "STR_SPLIT": exp.Split.from_arg_list,
286            "STR_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
287            "TO_TIMESTAMP": exp.UnixToTime.from_arg_list,
288            "UNNEST": exp.Explode.from_arg_list,
289            "XOR": binary_from_function(exp.BitwiseXor),
290            "GENERATE_SERIES": _build_generate_series(),
291            "RANGE": _build_generate_series(end_exclusive=True),
292        }
293
294        FUNCTION_PARSERS = parser.Parser.FUNCTION_PARSERS.copy()
295        FUNCTION_PARSERS.pop("DECODE")
296
297        NO_PAREN_FUNCTION_PARSERS = {
298            **parser.Parser.NO_PAREN_FUNCTION_PARSERS,
299            "MAP": lambda self: self._parse_map(),
300        }
301
302        TABLE_ALIAS_TOKENS = parser.Parser.TABLE_ALIAS_TOKENS - {
303            TokenType.SEMI,
304            TokenType.ANTI,
305        }
306
307        PLACEHOLDER_PARSERS = {
308            **parser.Parser.PLACEHOLDER_PARSERS,
309            TokenType.PARAMETER: lambda self: (
310                self.expression(exp.Placeholder, this=self._prev.text)
311                if self._match(TokenType.NUMBER) or self._match_set(self.ID_VAR_TOKENS)
312                else None
313            ),
314        }
315
316        def _parse_bracket(
317            self, this: t.Optional[exp.Expression] = None
318        ) -> t.Optional[exp.Expression]:
319            bracket = super()._parse_bracket(this)
320            if isinstance(bracket, exp.Bracket):
321                bracket.set("returns_list_for_maps", True)
322
323            return bracket
324
325        def _parse_map(self) -> exp.ToMap | exp.Map:
326            if self._match(TokenType.L_BRACE, advance=False):
327                return self.expression(exp.ToMap, this=self._parse_bracket())
328
329            args = self._parse_wrapped_csv(self._parse_conjunction)
330            return self.expression(exp.Map, keys=seq_get(args, 0), values=seq_get(args, 1))
331
332        def _parse_types(
333            self, check_func: bool = False, schema: bool = False, allow_identifiers: bool = True
334        ) -> t.Optional[exp.Expression]:
335            this = super()._parse_types(
336                check_func=check_func, schema=schema, allow_identifiers=allow_identifiers
337            )
338
339            # DuckDB treats NUMERIC and DECIMAL without precision as DECIMAL(18, 3)
340            # See: https://duckdb.org/docs/sql/data_types/numeric
341            if (
342                isinstance(this, exp.DataType)
343                and this.is_type("numeric", "decimal")
344                and not this.expressions
345            ):
346                return exp.DataType.build("DECIMAL(18, 3)")
347
348            return this
349
350        def _parse_struct_types(self, type_required: bool = False) -> t.Optional[exp.Expression]:
351            return self._parse_field_def()
352
353        def _pivot_column_names(self, aggregations: t.List[exp.Expression]) -> t.List[str]:
354            if len(aggregations) == 1:
355                return super()._pivot_column_names(aggregations)
356            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', 'STRUCT_PACK'}
FUNCTIONS = {'ABS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Abs'>>, 'ADD_MONTHS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.AddMonths'>>, 'ANONYMOUS_AGG_FUNC': <bound method Func.from_arg_list of <class 'sqlglot.expressions.AnonymousAggFunc'>>, 'ANY_VALUE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.AnyValue'>>, 'APPROX_DISTINCT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ApproxDistinct'>>, 'APPROX_COUNT_DISTINCT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ApproxDistinct'>>, 'APPROX_QUANTILE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ApproxQuantile'>>, 'APPROX_TOP_K': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ApproxTopK'>>, 'ARG_MAX': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArgMax'>>, 'ARGMAX': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArgMax'>>, 'MAX_BY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArgMax'>>, 'ARG_MIN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArgMin'>>, 'ARGMIN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArgMin'>>, 'MIN_BY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArgMin'>>, 'ARRAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Array'>>, 'ARRAY_AGG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayAgg'>>, 'ARRAY_ALL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayAll'>>, 'ARRAY_ANY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayAny'>>, 'ARRAY_CONCAT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayConcat'>>, 'ARRAY_CAT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayConcat'>>, 'ARRAY_CONTAINS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayContains'>>, 'FILTER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayFilter'>>, 'ARRAY_FILTER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayFilter'>>, 'ARRAY_OVERLAPS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayOverlaps'>>, 'ARRAY_SIZE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArraySize'>>, 'ARRAY_LENGTH': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArraySize'>>, 'ARRAY_SORT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.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_SUB': <bound method Func.from_arg_list of <class 'sqlglot.expressions.DateSub'>>, '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_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'>>, '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': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Hex'>>, 'HLL': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Hll'>>, 'IF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.If'>>, 'IIF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.If'>>, 'INITCAP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Initcap'>>, 'IS_INF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.IsInf'>>, 'ISINF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.IsInf'>>, 'IS_NAN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.IsNan'>>, 'ISNAN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.IsNan'>>, 'J_S_O_N_ARRAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONArray'>>, 'J_S_O_N_ARRAY_AGG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONArrayAgg'>>, 'JSON_ARRAY_CONTAINS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONArrayContains'>>, 'JSONB_EXTRACT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONBExtract'>>, 'JSONB_EXTRACT_SCALAR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.JSONBExtractScalar'>>, 'JSON_EXTRACT': <function build_extract_json_with_path.<locals>._builder>, 'JSON_EXTRACT_SCALAR': <function 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'>>, 'LN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Ln'>>, 'LOG': <function build_logarithm>, 'LOGICAL_AND': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LogicalAnd'>>, 'BOOL_AND': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LogicalAnd'>>, 'BOOLAND_AGG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LogicalAnd'>>, 'LOGICAL_OR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LogicalOr'>>, 'BOOL_OR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LogicalOr'>>, 'BOOLOR_AGG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.LogicalOr'>>, 'LOWER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Lower'>>, 'LCASE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Lower'>>, 'MD5': <bound method Func.from_arg_list of <class 'sqlglot.expressions.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'>>, 'OPEN_J_S_O_N': <bound method Func.from_arg_list of <class 'sqlglot.expressions.OpenJSON'>>, 'PARAMETERIZED_AGG': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ParameterizedAgg'>>, 'PARSE_JSON': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ParseJSON'>>, 'JSON_PARSE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ParseJSON'>>, 'PERCENTILE_CONT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.PercentileCont'>>, 'PERCENTILE_DISC': <bound method Func.from_arg_list of <class 'sqlglot.expressions.PercentileDisc'>>, 'POSEXPLODE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Posexplode'>>, 'POSEXPLODE_OUTER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.PosexplodeOuter'>>, 'POWER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Pow'>>, 'POW': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Pow'>>, 'PREDICT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Predict'>>, 'QUANTILE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Quantile'>>, 'RAND': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Rand'>>, 'RANDOM': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Rand'>>, 'RANDN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Randn'>>, 'RANGE_N': <bound method Func.from_arg_list of <class 'sqlglot.expressions.RangeN'>>, 'READ_CSV': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ReadCSV'>>, 'REDUCE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Reduce'>>, 'REGEXP_EXTRACT': <function 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'>>, 'STDDEV_POP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StddevPop'>>, 'STDDEV_SAMP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StddevSamp'>>, 'STR_POSITION': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StrPosition'>>, 'STR_TO_DATE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StrToDate'>>, 'STR_TO_MAP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StrToMap'>>, 'STR_TO_TIME': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StrToTime'>>, 'STR_TO_UNIX': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StrToUnix'>>, 'STRUCT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Struct'>>, 'STRUCT_EXTRACT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.StructExtract'>>, 'STUFF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Stuff'>>, 'INSERT': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Stuff'>>, 'SUBSTRING': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Substring'>>, 'SUM': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Sum'>>, 'TIME_ADD': <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_CAST': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TryCast'>>, 'TS_OR_DI_TO_DI': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDiToDi'>>, 'TS_OR_DS_ADD': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDsAdd'>>, 'TS_OR_DS_DIFF': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDsDiff'>>, 'TS_OR_DS_TO_DATE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDsToDate'>>, 'TS_OR_DS_TO_DATE_STR': <function Parser.<lambda>>, 'TS_OR_DS_TO_TIME': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDsToTime'>>, 'TS_OR_DS_TO_TIMESTAMP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.TsOrDsToTimestamp'>>, 'UNHEX': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Unhex'>>, 'UNIX_DATE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.UnixDate'>>, 'UNIX_TO_STR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.UnixToStr'>>, 'UNIX_TO_TIME': <bound method Func.from_arg_list of <class 'sqlglot.expressions.UnixToTime'>>, 'UNIX_TO_TIME_STR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.UnixToTimeStr'>>, 'UPPER': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Upper'>>, 'UCASE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Upper'>>, 'VAR_MAP': <function build_var_map>, 'VARIANCE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Variance'>>, 'VARIANCE_SAMP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Variance'>>, 'VAR_SAMP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Variance'>>, 'VARIANCE_POP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.VariancePop'>>, 'VAR_POP': <bound method Func.from_arg_list of <class 'sqlglot.expressions.VariancePop'>>, 'WEEK': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Week'>>, 'WEEK_OF_YEAR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.WeekOfYear'>>, 'WEEKOFYEAR': <bound method Func.from_arg_list of <class 'sqlglot.expressions.WeekOfYear'>>, 'WHEN': <bound method Func.from_arg_list of <class 'sqlglot.expressions.When'>>, 'X_M_L_TABLE': <bound method Func.from_arg_list of <class 'sqlglot.expressions.XMLTable'>>, 'XOR': <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>>, 'MOD': <function Parser.<lambda>>, 'ARRAY_HAS': <bound method Func.from_arg_list of <class 'sqlglot.expressions.ArrayContains'>>, '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'>>, 'STRING_TO_ARRAY': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Split'>>, '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'>>, 'UNNEST': <bound method Func.from_arg_list of <class 'sqlglot.expressions.Explode'>>, 'RANGE': <function _build_generate_series.<locals>._builder>}
FUNCTION_PARSERS = {'CAST': <function Parser.<lambda>>, 'CONVERT': <function Parser.<lambda>>, 'EXTRACT': <function Parser.<lambda>>, 'JSON_OBJECT': <function Parser.<lambda>>, 'JSON_OBJECTAGG': <function Parser.<lambda>>, 'JSON_TABLE': <function Parser.<lambda>>, 'MATCH': <function Parser.<lambda>>, 'OPENJSON': <function Parser.<lambda>>, 'POSITION': <function Parser.<lambda>>, 'PREDICT': <function Parser.<lambda>>, 'SAFE_CAST': <function Parser.<lambda>>, 'STRING_AGG': <function Parser.<lambda>>, 'SUBSTRING': <function Parser.<lambda>>, '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>>, 'IF': <function Parser.<lambda>>, 'NEXT': <function Parser.<lambda>>, 'MAP': <function DuckDB.Parser.<lambda>>}
TABLE_ALIAS_TOKENS = {<TokenType.SETTINGS: 'SETTINGS'>, <TokenType.ISNULL: 'ISNULL'>, <TokenType.TABLE: 'TABLE'>, <TokenType.JSON: 'JSON'>, <TokenType.TEXT: 'TEXT'>, <TokenType.BINARY: 'BINARY'>, <TokenType.COLLATE: 'COLLATE'>, <TokenType.GEOGRAPHY: 'GEOGRAPHY'>, <TokenType.DATERANGE: 'DATERANGE'>, <TokenType.COMMIT: 'COMMIT'>, <TokenType.FOREIGN_KEY: 'FOREIGN_KEY'>, <TokenType.IPV6: 'IPV6'>, <TokenType.MONEY: 'MONEY'>, <TokenType.LONGTEXT: 'LONGTEXT'>, <TokenType.INT4MULTIRANGE: 'INT4MULTIRANGE'>, <TokenType.UUID: 'UUID'>, <TokenType.FALSE: 'FALSE'>, <TokenType.SMALLSERIAL: 'SMALLSERIAL'>, <TokenType.EXISTS: 'EXISTS'>, <TokenType.PROCEDURE: 'PROCEDURE'>, <TokenType.PRAGMA: 'PRAGMA'>, <TokenType.COMMAND: 'COMMAND'>, <TokenType.YEAR: 'YEAR'>, <TokenType.IS: 'IS'>, <TokenType.UNPIVOT: 'UNPIVOT'>, <TokenType.ENUM16: 'ENUM16'>, <TokenType.INT256: 'INT256'>, <TokenType.FINAL: 'FINAL'>, <TokenType.SERIAL: 'SERIAL'>, <TokenType.MERGE: 'MERGE'>, <TokenType.ORDINALITY: 'ORDINALITY'>, <TokenType.NULLABLE: 'NULLABLE'>, <TokenType.IPPREFIX: 'IPPREFIX'>, <TokenType.STRUCT: 'STRUCT'>, <TokenType.STORAGE_INTEGRATION: 'STORAGE_INTEGRATION'>, <TokenType.NCHAR: 'NCHAR'>, <TokenType.TINYINT: 'TINYINT'>, <TokenType.REPLACE: 'REPLACE'>, <TokenType.UINT128: 'UINT128'>, <TokenType.AGGREGATEFUNCTION: 'AGGREGATEFUNCTION'>, <TokenType.NULL: 'NULL'>, <TokenType.SHOW: 'SHOW'>, <TokenType.REFERENCES: 'REFERENCES'>, <TokenType.FUNCTION: 'FUNCTION'>, <TokenType.USMALLINT: 'USMALLINT'>, <TokenType.RECURSIVE: 'RECURSIVE'>, <TokenType.INDEX: 'INDEX'>, <TokenType.FORMAT: 'FORMAT'>, <TokenType.RANGE: 'RANGE'>, <TokenType.UNIQUE: 'UNIQUE'>, <TokenType.OBJECT: 'OBJECT'>, <TokenType.END: 'END'>, <TokenType.USERDEFINED: 'USERDEFINED'>, <TokenType.UDECIMAL: 'UDECIMAL'>, <TokenType.TINYBLOB: 'TINYBLOB'>, <TokenType.VARIANT: 'VARIANT'>, <TokenType.NUMMULTIRANGE: 'NUMMULTIRANGE'>, <TokenType.OBJECT_IDENTIFIER: 'OBJECT_IDENTIFIER'>, <TokenType.NVARCHAR: 'NVARCHAR'>, <TokenType.BIGSERIAL: 'BIGSERIAL'>, <TokenType.INET: 'INET'>, <TokenType.NEXT: 'NEXT'>, <TokenType.DATABASE: 'DATABASE'>, <TokenType.OPERATOR: 'OPERATOR'>, <TokenType.OVERWRITE: 'OVERWRITE'>, <TokenType.CASE: 'CASE'>, <TokenType.UTINYINT: 'UTINYINT'>, <TokenType.LOWCARDINALITY: 'LOWCARDINALITY'>, <TokenType.CURRENT_TIME: 'CURRENT_TIME'>, <TokenType.REFRESH: 'REFRESH'>, <TokenType.BIGDECIMAL: 'BIGDECIMAL'>, <TokenType.EXECUTE: 'EXECUTE'>, <TokenType.CURRENT_TIMESTAMP: 'CURRENT_TIMESTAMP'>, <TokenType.TSMULTIRANGE: 'TSMULTIRANGE'>, <TokenType.BEGIN: 'BEGIN'>, <TokenType.SOME: 'SOME'>, <TokenType.LOAD: 'LOAD'>, <TokenType.TSTZMULTIRANGE: 'TSTZMULTIRANGE'>, <TokenType.DATE: 'DATE'>, <TokenType.COLUMN: 'COLUMN'>, <TokenType.DELETE: 'DELETE'>, <TokenType.DESCRIBE: 'DESCRIBE'>, <TokenType.TOP: 'TOP'>, <TokenType.TIME: 'TIME'>, <TokenType.BIT: 'BIT'>, <TokenType.DATETIME64: 'DATETIME64'>, <TokenType.DOUBLE: 'DOUBLE'>, <TokenType.DEFAULT: 'DEFAULT'>, <TokenType.SCHEMA: 'SCHEMA'>, <TokenType.DATEMULTIRANGE: 'DATEMULTIRANGE'>, <TokenType.FILTER: 'FILTER'>, <TokenType.HSTORE: 'HSTORE'>, <TokenType.PSEUDO_TYPE: 'PSEUDO_TYPE'>, <TokenType.TIMESTAMPLTZ: 'TIMESTAMPLTZ'>, <TokenType.UNKNOWN: 'UNKNOWN'>, <TokenType.GEOMETRY: 'GEOMETRY'>, <TokenType.KEEP: 'KEEP'>, <TokenType.IPV4: 'IPV4'>, <TokenType.IDENTIFIER: 'IDENTIFIER'>, <TokenType.TIMESTAMPTZ: 'TIMESTAMPTZ'>, <TokenType.PARTITION: 'PARTITION'>, <TokenType.TRUNCATE: 'TRUNCATE'>, <TokenType.DIV: 'DIV'>, <TokenType.FLOAT: 'FLOAT'>, <TokenType.XML: 'XML'>, <TokenType.TSTZRANGE: 'TSTZRANGE'>, <TokenType.MODEL: 'MODEL'>, <TokenType.MEDIUMINT: 'MEDIUMINT'>, <TokenType.NESTED: 'NESTED'>, <TokenType.ENUM: 'ENUM'>, <TokenType.INT128: 'INT128'>, <TokenType.INT8RANGE: 'INT8RANGE'>, <TokenType.COMMENT: 'COMMENT'>, <TokenType.DATE32: 'DATE32'>, <TokenType.ESCAPE: 'ESCAPE'>, <TokenType.FIXEDSTRING: 'FIXEDSTRING'>, <TokenType.USE: 'USE'>, <TokenType.DATETIME: 'DATETIME'>, <TokenType.TSRANGE: 'TSRANGE'>, <TokenType.KILL: 'KILL'>, <TokenType.VARCHAR: 'VARCHAR'>, <TokenType.MEDIUMTEXT: 'MEDIUMTEXT'>, <TokenType.NAME: 'NAME'>, <TokenType.FIRST: 'FIRST'>, <TokenType.ROW: 'ROW'>, <TokenType.SIMPLEAGGREGATEFUNCTION: 'SIMPLEAGGREGATEFUNCTION'>, <TokenType.TINYTEXT: 'TINYTEXT'>, <TokenType.MAP: 'MAP'>, <TokenType.VAR: 'VAR'>, <TokenType.INTERVAL: 'INTERVAL'>, <TokenType.OVERLAPS: 'OVERLAPS'>, <TokenType.CACHE: 'CACHE'>, <TokenType.INT4RANGE: 'INT4RANGE'>, <TokenType.UBIGINT: 'UBIGINT'>, <TokenType.ROWVERSION: 'ROWVERSION'>, <TokenType.BOOLEAN: 'BOOLEAN'>, <TokenType.AUTO_INCREMENT: 'AUTO_INCREMENT'>, <TokenType.PIVOT: 'PIVOT'>, <TokenType.CURRENT_DATETIME: 'CURRENT_DATETIME'>, <TokenType.BPCHAR: 'BPCHAR'>, <TokenType.VOLATILE: 'VOLATILE'>, <TokenType.ALL: 'ALL'>, <TokenType.IMAGE: 'IMAGE'>, <TokenType.UINT: 'UINT'>, <TokenType.CONSTRAINT: 'CONSTRAINT'>, <TokenType.TIMETZ: 'TIMETZ'>, <TokenType.DECIMAL: 'DECIMAL'>, <TokenType.SET: 'SET'>, <TokenType.UPDATE: 'UPDATE'>, <TokenType.LONGBLOB: 'LONGBLOB'>, <TokenType.INT8MULTIRANGE: 'INT8MULTIRANGE'>, <TokenType.ARRAY: 'ARRAY'>, <TokenType.IPADDRESS: 'IPADDRESS'>, <TokenType.PERCENT: 'PERCENT'>, <TokenType.CURRENT_USER: 'CURRENT_USER'>, <TokenType.CHAR: 'CHAR'>, <TokenType.TIMESTAMP_S: 'TIMESTAMP_S'>, <TokenType.UINT256: 'UINT256'>, <TokenType.TIMESTAMP_MS: 'TIMESTAMP_MS'>, <TokenType.ASC: 'ASC'>, <TokenType.TIMESTAMP: 'TIMESTAMP'>, <TokenType.VIEW: 'VIEW'>, <TokenType.MEDIUMBLOB: 'MEDIUMBLOB'>, <TokenType.HLLSKETCH: 'HLLSKETCH'>, <TokenType.SEQUENCE: 'SEQUENCE'>, <TokenType.DICTIONARY: 'DICTIONARY'>, <TokenType.JSONB: 'JSONB'>, <TokenType.ENUM8: 'ENUM8'>, <TokenType.UNIQUEIDENTIFIER: 'UNIQUEIDENTIFIER'>, <TokenType.BIGINT: 'BIGINT'>, <TokenType.UMEDIUMINT: 'UMEDIUMINT'>, <TokenType.ROWS: 'ROWS'>, <TokenType.TIMESTAMP_NS: 'TIMESTAMP_NS'>, <TokenType.INT: 'INT'>, <TokenType.TEMPORARY: 'TEMPORARY'>, <TokenType.SUPER: 'SUPER'>, <TokenType.SMALLMONEY: 'SMALLMONEY'>, <TokenType.SMALLINT: 'SMALLINT'>, <TokenType.ANY: 'ANY'>, <TokenType.VARBINARY: 'VARBINARY'>, <TokenType.NUMRANGE: 'NUMRANGE'>, <TokenType.CURRENT_DATE: 'CURRENT_DATE'>, <TokenType.DESC: 'DESC'>, <TokenType.TRUE: 'TRUE'>}
PLACEHOLDER_PARSERS = {<TokenType.PLACEHOLDER: 'PLACEHOLDER'>: <function Parser.<lambda>>, <TokenType.PARAMETER: 'PARAMETER'>: <function DuckDB.Parser.<lambda>>, <TokenType.COLON: 'COLON'>: <function Parser.<lambda>>}
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
ID_VAR_TOKENS
INTERVAL_VARS
ALIAS_TOKENS
COMMENT_TABLE_ALIAS_TOKENS
UPDATE_ALIAS_TOKENS
TRIM_TYPES
FUNC_TOKENS
CONJUNCTION
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
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
INSERT_ALTERNATIVES
CLONE_KEYWORDS
HISTORICAL_DATA_KIND
OPCLASS_FOLLOW_KEYWORDS
OPTYPE_FOLLOW_TOKENS
TABLE_INDEX_HINT_TOKENS
VIEW_ATTRIBUTES
WINDOW_ALIAS_TOKENS
WINDOW_BEFORE_PAREN_TOKENS
WINDOW_SIDES
JSON_KEY_VALUE_SEPARATOR_TOKENS
FETCH_TOKENS
ADD_CONSTRAINT_TOKENS
DISTINCT_TOKENS
NULL_TOKENS
UNNEST_OFFSET_ALIAS_TOKENS
SELECT_START_TOKENS
STRICT_CAST
PREFIXED_PIVOT_COLUMNS
IDENTIFY_PIVOT_STRINGS
LOG_DEFAULTS_TO_LN
ALTER_TABLE_ADD_REQUIRED_FOR_EACH_COLUMN
TABLESAMPLE_CSV
SET_REQUIRES_ASSIGNMENT_DELIMITER
TRIM_PATTERN_FIRST
STRING_ALIASES
MODIFIERS_ATTACHED_TO_UNION
UNION_MODIFIERS
NO_PAREN_IF_COMMANDS
JSON_ARROWS_REQUIRE_JSON_TYPE
VALUES_FOLLOWED_BY_PAREN
SUPPORTS_IMPLICIT_UNNEST
INTERVAL_SPANS
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):
358    class Generator(generator.Generator):
359        PARAMETER_TOKEN = "$"
360        NAMED_PLACEHOLDER_TOKEN = "$"
361        JOIN_HINTS = False
362        TABLE_HINTS = False
363        QUERY_HINTS = False
364        LIMIT_FETCH = "LIMIT"
365        STRUCT_DELIMITER = ("(", ")")
366        RENAME_TABLE_WITH_DB = False
367        NVL2_SUPPORTED = False
368        SEMI_ANTI_JOIN_WITH_SIDE = False
369        TABLESAMPLE_KEYWORDS = "USING SAMPLE"
370        TABLESAMPLE_SEED_KEYWORD = "REPEATABLE"
371        LAST_DAY_SUPPORTS_DATE_PART = False
372        JSON_KEY_VALUE_PAIR_SEP = ","
373        IGNORE_NULLS_IN_FUNC = True
374        JSON_PATH_BRACKETED_KEY_SUPPORTED = False
375        SUPPORTS_CREATE_TABLE_LIKE = False
376        MULTI_ARG_DISTINCT = False
377        CAN_IMPLEMENT_ARRAY_ANY = True
378        SUPPORTS_TO_NUMBER = False
379
380        TRANSFORMS = {
381            **generator.Generator.TRANSFORMS,
382            exp.ApproxDistinct: approx_count_distinct_sql,
383            exp.Array: inline_array_unless_query,
384            exp.ArrayFilter: rename_func("LIST_FILTER"),
385            exp.ArraySize: rename_func("ARRAY_LENGTH"),
386            exp.ArgMax: arg_max_or_min_no_count("ARG_MAX"),
387            exp.ArgMin: arg_max_or_min_no_count("ARG_MIN"),
388            exp.ArraySort: _array_sort_sql,
389            exp.ArraySum: rename_func("LIST_SUM"),
390            exp.BitwiseXor: rename_func("XOR"),
391            exp.CommentColumnConstraint: no_comment_column_constraint_sql,
392            exp.CurrentDate: lambda *_: "CURRENT_DATE",
393            exp.CurrentTime: lambda *_: "CURRENT_TIME",
394            exp.CurrentTimestamp: lambda *_: "CURRENT_TIMESTAMP",
395            exp.DayOfMonth: rename_func("DAYOFMONTH"),
396            exp.DayOfWeek: rename_func("DAYOFWEEK"),
397            exp.DayOfYear: rename_func("DAYOFYEAR"),
398            exp.DataType: _datatype_sql,
399            exp.Date: _date_sql,
400            exp.DateAdd: _date_delta_sql,
401            exp.DateFromParts: rename_func("MAKE_DATE"),
402            exp.DateSub: _date_delta_sql,
403            exp.DateDiff: lambda self, e: self.func(
404                "DATE_DIFF", f"'{e.args.get('unit') or 'DAY'}'", e.expression, e.this
405            ),
406            exp.DateStrToDate: datestrtodate_sql,
407            exp.DateToDi: lambda self,
408            e: f"CAST(STRFTIME({self.sql(e, 'this')}, {DuckDB.DATEINT_FORMAT}) AS INT)",
409            exp.Decode: lambda self, e: encode_decode_sql(self, e, "DECODE", replace=False),
410            exp.DiToDate: lambda self,
411            e: f"CAST(STRPTIME(CAST({self.sql(e, 'this')} AS TEXT), {DuckDB.DATEINT_FORMAT}) AS DATE)",
412            exp.Encode: lambda self, e: encode_decode_sql(self, e, "ENCODE", replace=False),
413            exp.Explode: rename_func("UNNEST"),
414            exp.IntDiv: lambda self, e: self.binary(e, "//"),
415            exp.IsInf: rename_func("ISINF"),
416            exp.IsNan: rename_func("ISNAN"),
417            exp.JSONExtract: arrow_json_extract_sql,
418            exp.JSONExtractScalar: arrow_json_extract_sql,
419            exp.JSONFormat: _json_format_sql,
420            exp.LogicalOr: rename_func("BOOL_OR"),
421            exp.LogicalAnd: rename_func("BOOL_AND"),
422            exp.MonthsBetween: lambda self, e: self.func(
423                "DATEDIFF",
424                "'month'",
425                exp.cast(e.expression, "timestamp", copy=True),
426                exp.cast(e.this, "timestamp", copy=True),
427            ),
428            exp.ParseJSON: rename_func("JSON"),
429            exp.PercentileCont: _rename_unless_within_group("PERCENTILE_CONT", "QUANTILE_CONT"),
430            exp.PercentileDisc: _rename_unless_within_group("PERCENTILE_DISC", "QUANTILE_DISC"),
431            # DuckDB doesn't allow qualified columns inside of PIVOT expressions.
432            # See: https://github.com/duckdb/duckdb/blob/671faf92411182f81dce42ac43de8bfb05d9909e/src/planner/binder/tableref/bind_pivot.cpp#L61-L62
433            exp.Pivot: transforms.preprocess([transforms.unqualify_columns]),
434            exp.RegexpExtract: regexp_extract_sql,
435            exp.RegexpReplace: lambda self, e: self.func(
436                "REGEXP_REPLACE",
437                e.this,
438                e.expression,
439                e.args.get("replacement"),
440                e.args.get("modifiers"),
441            ),
442            exp.RegexpLike: rename_func("REGEXP_MATCHES"),
443            exp.RegexpSplit: rename_func("STR_SPLIT_REGEX"),
444            exp.Rand: rename_func("RANDOM"),
445            exp.SafeDivide: no_safe_divide_sql,
446            exp.Split: rename_func("STR_SPLIT"),
447            exp.SortArray: _sort_array_sql,
448            exp.StrPosition: str_position_sql,
449            exp.StrToDate: lambda self, e: f"CAST({str_to_time_sql(self, e)} AS DATE)",
450            exp.StrToTime: str_to_time_sql,
451            exp.StrToUnix: lambda self, e: self.func(
452                "EPOCH", self.func("STRPTIME", e.this, self.format_time(e))
453            ),
454            exp.Struct: _struct_sql,
455            exp.TimeAdd: _date_delta_sql,
456            exp.Timestamp: no_timestamp_sql,
457            exp.TimestampDiff: lambda self, e: self.func(
458                "DATE_DIFF", exp.Literal.string(e.unit), e.expression, e.this
459            ),
460            exp.TimestampTrunc: timestamptrunc_sql,
461            exp.TimeStrToDate: lambda self, e: self.sql(exp.cast(e.this, "date")),
462            exp.TimeStrToTime: timestrtotime_sql,
463            exp.TimeStrToUnix: lambda self, e: self.func("EPOCH", exp.cast(e.this, "timestamp")),
464            exp.TimeToStr: lambda self, e: self.func("STRFTIME", e.this, self.format_time(e)),
465            exp.TimeToUnix: rename_func("EPOCH"),
466            exp.TsOrDiToDi: lambda self,
467            e: f"CAST(SUBSTR(REPLACE(CAST({self.sql(e, 'this')} AS TEXT), '-', ''), 1, 8) AS INT)",
468            exp.TsOrDsAdd: _ts_or_ds_add_sql,
469            exp.TsOrDsDiff: lambda self, e: self.func(
470                "DATE_DIFF",
471                f"'{e.args.get('unit') or 'DAY'}'",
472                exp.cast(e.expression, "TIMESTAMP"),
473                exp.cast(e.this, "TIMESTAMP"),
474            ),
475            exp.UnixToStr: lambda self, e: self.func(
476                "STRFTIME", self.func("TO_TIMESTAMP", e.this), self.format_time(e)
477            ),
478            exp.UnixToTime: _unix_to_time_sql,
479            exp.UnixToTimeStr: lambda self, e: f"CAST(TO_TIMESTAMP({self.sql(e, 'this')}) AS TEXT)",
480            exp.VariancePop: rename_func("VAR_POP"),
481            exp.WeekOfYear: rename_func("WEEKOFYEAR"),
482            exp.Xor: bool_xor_sql,
483        }
484
485        SUPPORTED_JSON_PATH_PARTS = {
486            exp.JSONPathKey,
487            exp.JSONPathRoot,
488            exp.JSONPathSubscript,
489            exp.JSONPathWildcard,
490        }
491
492        TYPE_MAPPING = {
493            **generator.Generator.TYPE_MAPPING,
494            exp.DataType.Type.BINARY: "BLOB",
495            exp.DataType.Type.CHAR: "TEXT",
496            exp.DataType.Type.FLOAT: "REAL",
497            exp.DataType.Type.NCHAR: "TEXT",
498            exp.DataType.Type.NVARCHAR: "TEXT",
499            exp.DataType.Type.UINT: "UINTEGER",
500            exp.DataType.Type.VARBINARY: "BLOB",
501            exp.DataType.Type.VARCHAR: "TEXT",
502            exp.DataType.Type.TIMESTAMP_S: "TIMESTAMP_S",
503            exp.DataType.Type.TIMESTAMP_MS: "TIMESTAMP_MS",
504            exp.DataType.Type.TIMESTAMP_NS: "TIMESTAMP_NS",
505        }
506
507        STAR_MAPPING = {**generator.Generator.STAR_MAPPING, "except": "EXCLUDE"}
508
509        UNWRAPPED_INTERVAL_VALUES = (exp.Literal, exp.Paren)
510
511        # DuckDB doesn't generally support CREATE TABLE .. properties
512        # https://duckdb.org/docs/sql/statements/create_table.html
513        PROPERTIES_LOCATION = {
514            prop: exp.Properties.Location.UNSUPPORTED
515            for prop in generator.Generator.PROPERTIES_LOCATION
516        }
517
518        # There are a few exceptions (e.g. temporary tables) which are supported or
519        # can be transpiled to DuckDB, so we explicitly override them accordingly
520        PROPERTIES_LOCATION[exp.LikeProperty] = exp.Properties.Location.POST_SCHEMA
521        PROPERTIES_LOCATION[exp.TemporaryProperty] = exp.Properties.Location.POST_CREATE
522
523        def timefromparts_sql(self, expression: exp.TimeFromParts) -> str:
524            nano = expression.args.get("nano")
525            if nano is not None:
526                expression.set(
527                    "sec", expression.args["sec"] + nano.pop() / exp.Literal.number(1000000000.0)
528                )
529
530            return rename_func("MAKE_TIME")(self, expression)
531
532        def timestampfromparts_sql(self, expression: exp.TimestampFromParts) -> str:
533            sec = expression.args["sec"]
534
535            milli = expression.args.get("milli")
536            if milli is not None:
537                sec += milli.pop() / exp.Literal.number(1000.0)
538
539            nano = expression.args.get("nano")
540            if nano is not None:
541                sec += nano.pop() / exp.Literal.number(1000000000.0)
542
543            if milli or nano:
544                expression.set("sec", sec)
545
546            return rename_func("MAKE_TIMESTAMP")(self, expression)
547
548        def tablesample_sql(
549            self,
550            expression: exp.TableSample,
551            sep: str = " AS ",
552            tablesample_keyword: t.Optional[str] = None,
553        ) -> str:
554            if not isinstance(expression.parent, exp.Select):
555                # This sample clause only applies to a single source, not the entire resulting relation
556                tablesample_keyword = "TABLESAMPLE"
557
558            return super().tablesample_sql(
559                expression, sep=sep, tablesample_keyword=tablesample_keyword
560            )
561
562        def interval_sql(self, expression: exp.Interval) -> str:
563            multiplier: t.Optional[int] = None
564            unit = expression.text("unit").lower()
565
566            if unit.startswith("week"):
567                multiplier = 7
568            if unit.startswith("quarter"):
569                multiplier = 90
570
571            if multiplier:
572                return f"({multiplier} * {super().interval_sql(exp.Interval(this=expression.this, unit=exp.var('DAY')))})"
573
574            return super().interval_sql(expression)
575
576        def columndef_sql(self, expression: exp.ColumnDef, sep: str = " ") -> str:
577            if isinstance(expression.parent, exp.UserDefinedFunction):
578                return self.sql(expression, "this")
579            return super().columndef_sql(expression, sep)
580
581        def join_sql(self, expression: exp.Join) -> str:
582            if (
583                expression.side == "LEFT"
584                and not expression.args.get("on")
585                and isinstance(expression.this, exp.Unnest)
586            ):
587                # Some dialects support `LEFT JOIN UNNEST(...)` without an explicit ON clause
588                # DuckDB doesn't, but we can just add a dummy ON clause that is always true
589                return super().join_sql(expression.on(exp.true()))
590
591            return super().join_sql(expression)
592
593        def generateseries_sql(self, expression: exp.GenerateSeries) -> str:
594            # GENERATE_SERIES(a, b) -> [a, b], RANGE(a, b) -> [a, b)
595            if expression.args.get("is_end_exclusive"):
596                expression.set("is_end_exclusive", None)
597                return rename_func("RANGE")(self, expression)
598
599            return super().generateseries_sql(expression)
600
601        def bracket_sql(self, expression: exp.Bracket) -> str:
602            this = expression.this
603            if isinstance(this, exp.Array):
604                this.replace(exp.paren(this))
605
606            bracket = super().bracket_sql(expression)
607
608            if not expression.args.get("returns_list_for_maps"):
609                if not this.type:
610                    from sqlglot.optimizer.annotate_types import annotate_types
611
612                    this = annotate_types(this)
613
614                if this.is_type(exp.DataType.Type.MAP):
615                    bracket = f"({bracket})[1]"
616
617            return bracket

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
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.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.CopyGrantsProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.DateFormatColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.DefaultColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.EncodeColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.EphemeralColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ExcludeColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ExecuteAsProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ExternalProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.GlobalProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.HeapProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.IcebergProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.InheritsProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.InlineLengthColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.InputModelProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.IntervalSpan'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.JSONExtract'>: <function arrow_json_extract_sql>, <class 'sqlglot.expressions.JSONExtractScalar'>: <function arrow_json_extract_sql>, <class 'sqlglot.expressions.LanguageProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.LocationProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.LogProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.MaterializedProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.NonClusteredColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.NoPrimaryIndexProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.NotForReplicationColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.OnCommitProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.OnProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.OnUpdateColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.OutputModelProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.PathColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.RemoteWithConnectionModelProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ReturnsProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SampleProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SetConfigProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SetProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SettingsProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SharingProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SqlReadWriteProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.SqlSecurityProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.StabilityProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.TemporaryProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.TitleColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.Timestamp'>: <function no_timestamp_sql>, <class 'sqlglot.expressions.ToMap'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ToTableProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.TransformModelProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.TransientProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.UppercaseColumnConstraint'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.UnloggedProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.VarMap'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ViewAttributeProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.VolatileProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.WithJournalTableProperty'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.WithOperator'>: <function Generator.<lambda>>, <class 'sqlglot.expressions.ApproxDistinct'>: <function approx_count_distinct_sql>, <class 'sqlglot.expressions.Array'>: <function inline_array_unless_query>, <class 'sqlglot.expressions.ArrayFilter'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.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.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.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.MonthsBetween'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.ParseJSON'>: <function rename_func.<locals>.<lambda>>, <class 'sqlglot.expressions.PercentileCont'>: <function _rename_unless_within_group.<locals>.<lambda>>, <class 'sqlglot.expressions.PercentileDisc'>: <function _rename_unless_within_group.<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.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.StrToDate'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.StrToTime'>: <function str_to_time_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.TimestampDiff'>: <function DuckDB.Generator.<lambda>>, <class 'sqlglot.expressions.TimestampTrunc'>: <function 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 _ts_or_ds_add_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.BINARY: 'BINARY'>: 'BLOB', <Type.CHAR: 'CHAR'>: 'TEXT', <Type.FLOAT: 'FLOAT'>: 'REAL', <Type.UINT: 'UINT'>: 'UINTEGER', <Type.VARBINARY: 'VARBINARY'>: 'BLOB', <Type.VARCHAR: 'VARCHAR'>: 'TEXT', <Type.TIMESTAMP_S: 'TIMESTAMP_S'>: 'TIMESTAMP_S', <Type.TIMESTAMP_MS: 'TIMESTAMP_MS'>: 'TIMESTAMP_MS', <Type.TIMESTAMP_NS: 'TIMESTAMP_NS'>: 'TIMESTAMP_NS'}
STAR_MAPPING = {'except': 'EXCLUDE', 'replace': 'REPLACE'}
UNWRAPPED_INTERVAL_VALUES = (<class 'sqlglot.expressions.Literal'>, <class 'sqlglot.expressions.Paren'>)
PROPERTIES_LOCATION = {<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.DefinerProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.DictRange'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>, <class 'sqlglot.expressions.DictProperty'>: <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.UNSUPPORTED: 'UNSUPPORTED'>, <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.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.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.WithSystemVersioningProperty'>: <Location.UNSUPPORTED: 'UNSUPPORTED'>}
def timefromparts_sql(self, expression: sqlglot.expressions.TimeFromParts) -> str:
523        def timefromparts_sql(self, expression: exp.TimeFromParts) -> str:
524            nano = expression.args.get("nano")
525            if nano is not None:
526                expression.set(
527                    "sec", expression.args["sec"] + nano.pop() / exp.Literal.number(1000000000.0)
528                )
529
530            return rename_func("MAKE_TIME")(self, expression)
def timestampfromparts_sql(self, expression: sqlglot.expressions.TimestampFromParts) -> str:
532        def timestampfromparts_sql(self, expression: exp.TimestampFromParts) -> str:
533            sec = expression.args["sec"]
534
535            milli = expression.args.get("milli")
536            if milli is not None:
537                sec += milli.pop() / exp.Literal.number(1000.0)
538
539            nano = expression.args.get("nano")
540            if nano is not None:
541                sec += nano.pop() / exp.Literal.number(1000000000.0)
542
543            if milli or nano:
544                expression.set("sec", sec)
545
546            return rename_func("MAKE_TIMESTAMP")(self, expression)
def tablesample_sql( self, expression: sqlglot.expressions.TableSample, sep: str = ' AS ', tablesample_keyword: Optional[str] = None) -> str:
548        def tablesample_sql(
549            self,
550            expression: exp.TableSample,
551            sep: str = " AS ",
552            tablesample_keyword: t.Optional[str] = None,
553        ) -> str:
554            if not isinstance(expression.parent, exp.Select):
555                # This sample clause only applies to a single source, not the entire resulting relation
556                tablesample_keyword = "TABLESAMPLE"
557
558            return super().tablesample_sql(
559                expression, sep=sep, tablesample_keyword=tablesample_keyword
560            )
def interval_sql(self, expression: sqlglot.expressions.Interval) -> str:
562        def interval_sql(self, expression: exp.Interval) -> str:
563            multiplier: t.Optional[int] = None
564            unit = expression.text("unit").lower()
565
566            if unit.startswith("week"):
567                multiplier = 7
568            if unit.startswith("quarter"):
569                multiplier = 90
570
571            if multiplier:
572                return f"({multiplier} * {super().interval_sql(exp.Interval(this=expression.this, unit=exp.var('DAY')))})"
573
574            return super().interval_sql(expression)
def columndef_sql(self, expression: sqlglot.expressions.ColumnDef, sep: str = ' ') -> str:
576        def columndef_sql(self, expression: exp.ColumnDef, sep: str = " ") -> str:
577            if isinstance(expression.parent, exp.UserDefinedFunction):
578                return self.sql(expression, "this")
579            return super().columndef_sql(expression, sep)
def join_sql(self, expression: sqlglot.expressions.Join) -> str:
581        def join_sql(self, expression: exp.Join) -> str:
582            if (
583                expression.side == "LEFT"
584                and not expression.args.get("on")
585                and isinstance(expression.this, exp.Unnest)
586            ):
587                # Some dialects support `LEFT JOIN UNNEST(...)` without an explicit ON clause
588                # DuckDB doesn't, but we can just add a dummy ON clause that is always true
589                return super().join_sql(expression.on(exp.true()))
590
591            return super().join_sql(expression)
def generateseries_sql(self, expression: sqlglot.expressions.GenerateSeries) -> str:
593        def generateseries_sql(self, expression: exp.GenerateSeries) -> str:
594            # GENERATE_SERIES(a, b) -> [a, b], RANGE(a, b) -> [a, b)
595            if expression.args.get("is_end_exclusive"):
596                expression.set("is_end_exclusive", None)
597                return rename_func("RANGE")(self, expression)
598
599            return super().generateseries_sql(expression)
def bracket_sql(self, expression: sqlglot.expressions.Bracket) -> str:
601        def bracket_sql(self, expression: exp.Bracket) -> str:
602            this = expression.this
603            if isinstance(this, exp.Array):
604                this.replace(exp.paren(this))
605
606            bracket = super().bracket_sql(expression)
607
608            if not expression.args.get("returns_list_for_maps"):
609                if not this.type:
610                    from sqlglot.optimizer.annotate_types import annotate_types
611
612                    this = annotate_types(this)
613
614                if this.is_type(exp.DataType.Type.MAP):
615                    bracket = f"({bracket})[1]"
616
617            return bracket
SELECT_KINDS: Tuple[str, ...] = ()
AFTER_HAVING_MODIFIER_TRANSFORMS = {'qualify': <function Generator.<lambda>>, 'windows': <function Generator.<lambda>>}
Inherited Members
sqlglot.generator.Generator
Generator
NULL_ORDERING_SUPPORTED
LOCKING_READS_SUPPORTED
EXPLICIT_UNION
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
COLUMN_JOIN_MARKS_SUPPORTED
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
TIME_PART_SINGULARS
TOKEN_MAPPING
RESERVED_KEYWORDS
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_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
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
queryoption_sql
offset_limit_modifiers
after_limit_modifiers
select_sql
schema_sql
schema_columns_sql
star_sql
parameter_sql
sessionparameter_sql
placeholder_sql
subquery_sql
qualify_sql
set_operations
union_sql
union_op
unnest_sql
prewhere_sql
where_sql
window_sql
partition_by_sql
windowspec_sql
withingroup_sql
between_sql
bracket_offset_expressions
all_sql
any_sql
exists_sql
case_sql
constraint_sql
nextvaluefor_sql
extract_sql
trim_sql
convert_concat_args
concat_sql
concatws_sql
check_sql
foreignkey_sql
primarykey_sql
if_sql
matchagainst_sql
jsonkeyvalue_sql
jsonpath_sql
json_path_part
formatjson_sql
jsonobject_sql
jsonobjectagg_sql
jsonarray_sql
jsonarrayagg_sql
jsoncolumndef_sql
jsonschema_sql
jsontable_sql
openjsoncolumndef_sql
openjson_sql
in_sql
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
currenttimestamp_sql
collate_sql
command_sql
comment_sql
mergetreettlaction_sql
mergetreettl_sql
transaction_sql
commit_sql
rollback_sql
altercolumn_sql
renametable_sql
renamecolumn_sql
altertable_sql
add_column_sql
droppartition_sql
addconstraint_sql
distinct_sql
ignorenulls_sql
respectnulls_sql
havingmax_sql
intdiv_sql
dpipe_sql
div_sql
overlaps_sql
distance_sql
dot_sql
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
log_sql
use_sql
binary
function_fallback_sql
func
format_args
text_width
format_time
expressions
op_expressions
naked_property
tag_sql
token_sql
userdefinedfunction_sql
joinhint_sql
kwarg_sql
when_sql
merge_sql
tochar_sql
tonumber_sql
dictproperty_sql
dictrange_sql
dictsubproperty_sql
oncluster_sql
clusteredbyproperty_sql
anyvalue_sql
querytransform_sql
indexconstraintoption_sql
checkcolumnconstraint_sql
indexcolumnconstraint_sql
nvl2_sql
comprehension_sql
columnprefix_sql
opclass_sql
predict_sql
forin_sql
refresh_sql
operator_sql
toarray_sql
tsordstotime_sql
tsordstotimestamp_sql
tsordstodate_sql
unixdate_sql
lastday_sql
dateadd_sql
arrayany_sql
struct_sql
partitionrange_sql
truncatetable_sql
convert_sql