Edit on GitHub

sqlglot.dialects.duckdb

  1from __future__ import annotations
  2
  3from sqlglot import exp, generator, parser, tokens
  4from sqlglot.dialects.dialect import (
  5    Dialect,
  6    approx_count_distinct_sql,
  7    arrow_json_extract_scalar_sql,
  8    arrow_json_extract_sql,
  9    datestrtodate_sql,
 10    format_time_lambda,
 11    no_pivot_sql,
 12    no_properties_sql,
 13    no_safe_divide_sql,
 14    rename_func,
 15    str_position_sql,
 16    str_to_time_sql,
 17    timestrtotime_sql,
 18    ts_or_ds_to_date_sql,
 19)
 20from sqlglot.helper import seq_get
 21from sqlglot.tokens import TokenType
 22
 23
 24def _ts_or_ds_add(self, expression):
 25    this = expression.args.get("this")
 26    unit = self.sql(expression, "unit").strip("'") or "DAY"
 27    return f"CAST({this} AS DATE) + {self.sql(exp.Interval(this=expression.expression, unit=unit))}"
 28
 29
 30def _date_add(self, expression):
 31    this = self.sql(expression, "this")
 32    unit = self.sql(expression, "unit").strip("'") or "DAY"
 33    return f"{this} + {self.sql(exp.Interval(this=expression.expression, unit=unit))}"
 34
 35
 36def _array_sort_sql(self, expression):
 37    if expression.expression:
 38        self.unsupported("DUCKDB ARRAY_SORT does not support a comparator")
 39    return f"ARRAY_SORT({self.sql(expression, 'this')})"
 40
 41
 42def _sort_array_sql(self, expression):
 43    this = self.sql(expression, "this")
 44    if expression.args.get("asc") == exp.false():
 45        return f"ARRAY_REVERSE_SORT({this})"
 46    return f"ARRAY_SORT({this})"
 47
 48
 49def _sort_array_reverse(args):
 50    return exp.SortArray(this=seq_get(args, 0), asc=exp.false())
 51
 52
 53def _struct_sql(self, expression):
 54    args = [
 55        f"'{e.name or e.this.name}': {self.sql(e, 'expression')}" for e in expression.expressions
 56    ]
 57    return f"{{{', '.join(args)}}}"
 58
 59
 60def _datatype_sql(self, expression):
 61    if expression.this == exp.DataType.Type.ARRAY:
 62        return f"{self.expressions(expression, flat=True)}[]"
 63    return self.datatype_sql(expression)
 64
 65
 66def _regexp_extract_sql(self, expression):
 67    bad_args = list(filter(expression.args.get, ("position", "occurrence")))
 68    if bad_args:
 69        self.unsupported(f"REGEXP_EXTRACT does not support arg(s) {bad_args}")
 70    return self.func(
 71        "REGEXP_EXTRACT",
 72        expression.args.get("this"),
 73        expression.args.get("expression"),
 74        expression.args.get("group"),
 75    )
 76
 77
 78class DuckDB(Dialect):
 79    class Tokenizer(tokens.Tokenizer):
 80        KEYWORDS = {
 81            **tokens.Tokenizer.KEYWORDS,
 82            ":=": TokenType.EQ,
 83            "ATTACH": TokenType.COMMAND,
 84            "BINARY": TokenType.VARBINARY,
 85            "BPCHAR": TokenType.TEXT,
 86            "BITSTRING": TokenType.BIT,
 87            "CHAR": TokenType.TEXT,
 88            "CHARACTER VARYING": TokenType.TEXT,
 89            "EXCLUDE": TokenType.EXCEPT,
 90            "INT1": TokenType.TINYINT,
 91            "LOGICAL": TokenType.BOOLEAN,
 92            "NUMERIC": TokenType.DOUBLE,
 93            "SIGNED": TokenType.INT,
 94            "STRING": TokenType.VARCHAR,
 95            "UBIGINT": TokenType.UBIGINT,
 96            "UINTEGER": TokenType.UINT,
 97            "USMALLINT": TokenType.USMALLINT,
 98            "UTINYINT": TokenType.UTINYINT,
 99        }
100
101    class Parser(parser.Parser):
102        FUNCTIONS = {
103            **parser.Parser.FUNCTIONS,  # type: ignore
104            "APPROX_COUNT_DISTINCT": exp.ApproxDistinct.from_arg_list,
105            "ARRAY_LENGTH": exp.ArraySize.from_arg_list,
106            "ARRAY_SORT": exp.SortArray.from_arg_list,
107            "ARRAY_REVERSE_SORT": _sort_array_reverse,
108            "EPOCH": exp.TimeToUnix.from_arg_list,
109            "EPOCH_MS": lambda args: exp.UnixToTime(
110                this=exp.Div(
111                    this=seq_get(args, 0),
112                    expression=exp.Literal.number(1000),
113                )
114            ),
115            "LIST_SORT": exp.SortArray.from_arg_list,
116            "LIST_REVERSE_SORT": _sort_array_reverse,
117            "LIST_VALUE": exp.Array.from_arg_list,
118            "REGEXP_MATCHES": exp.RegexpLike.from_arg_list,
119            "STRFTIME": format_time_lambda(exp.TimeToStr, "duckdb"),
120            "STRPTIME": format_time_lambda(exp.StrToTime, "duckdb"),
121            "STR_SPLIT": exp.Split.from_arg_list,
122            "STRING_SPLIT": exp.Split.from_arg_list,
123            "STRING_TO_ARRAY": exp.Split.from_arg_list,
124            "STR_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
125            "STRING_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
126            "STRUCT_PACK": exp.Struct.from_arg_list,
127            "TO_TIMESTAMP": exp.UnixToTime.from_arg_list,
128            "UNNEST": exp.Explode.from_arg_list,
129        }
130
131        TYPE_TOKENS = {
132            *parser.Parser.TYPE_TOKENS,
133            TokenType.UBIGINT,
134            TokenType.UINT,
135            TokenType.USMALLINT,
136            TokenType.UTINYINT,
137        }
138
139    class Generator(generator.Generator):
140        STRUCT_DELIMITER = ("(", ")")
141
142        TRANSFORMS = {
143            **generator.Generator.TRANSFORMS,  # type: ignore
144            exp.ApproxDistinct: approx_count_distinct_sql,
145            exp.Array: lambda self, e: self.func("ARRAY", e.expressions[0])
146            if isinstance(seq_get(e.expressions, 0), exp.Select)
147            else rename_func("LIST_VALUE")(self, e),
148            exp.ArraySize: rename_func("ARRAY_LENGTH"),
149            exp.ArraySort: _array_sort_sql,
150            exp.ArraySum: rename_func("LIST_SUM"),
151            exp.DataType: _datatype_sql,
152            exp.DateAdd: _date_add,
153            exp.DateDiff: lambda self, e: self.func(
154                "DATE_DIFF", e.args.get("unit") or exp.Literal.string("day"), e.expression, e.this
155            ),
156            exp.DateStrToDate: datestrtodate_sql,
157            exp.DateToDi: lambda self, e: f"CAST(STRFTIME({self.sql(e, 'this')}, {DuckDB.dateint_format}) AS INT)",
158            exp.DiToDate: lambda self, e: f"CAST(STRPTIME(CAST({self.sql(e, 'this')} AS TEXT), {DuckDB.dateint_format}) AS DATE)",
159            exp.Explode: rename_func("UNNEST"),
160            exp.JSONExtract: arrow_json_extract_sql,
161            exp.JSONExtractScalar: arrow_json_extract_scalar_sql,
162            exp.JSONBExtract: arrow_json_extract_sql,
163            exp.JSONBExtractScalar: arrow_json_extract_scalar_sql,
164            exp.LogicalOr: rename_func("BOOL_OR"),
165            exp.Pivot: no_pivot_sql,
166            exp.Properties: no_properties_sql,
167            exp.RegexpExtract: _regexp_extract_sql,
168            exp.RegexpLike: rename_func("REGEXP_MATCHES"),
169            exp.RegexpSplit: rename_func("STR_SPLIT_REGEX"),
170            exp.SafeDivide: no_safe_divide_sql,
171            exp.Split: rename_func("STR_SPLIT"),
172            exp.SortArray: _sort_array_sql,
173            exp.StrPosition: str_position_sql,
174            exp.StrToDate: lambda self, e: f"CAST({str_to_time_sql(self, e)} AS DATE)",
175            exp.StrToTime: str_to_time_sql,
176            exp.StrToUnix: lambda self, e: f"EPOCH(STRPTIME({self.sql(e, 'this')}, {self.format_time(e)}))",
177            exp.Struct: _struct_sql,
178            exp.TimeStrToDate: lambda self, e: f"CAST({self.sql(e, 'this')} AS DATE)",
179            exp.TimeStrToTime: timestrtotime_sql,
180            exp.TimeStrToUnix: lambda self, e: f"EPOCH(CAST({self.sql(e, 'this')} AS TIMESTAMP))",
181            exp.TimeToStr: lambda self, e: f"STRFTIME({self.sql(e, 'this')}, {self.format_time(e)})",
182            exp.TimeToUnix: rename_func("EPOCH"),
183            exp.TsOrDiToDi: lambda self, e: f"CAST(SUBSTR(REPLACE(CAST({self.sql(e, 'this')} AS TEXT), '-', ''), 1, 8) AS INT)",
184            exp.TsOrDsAdd: _ts_or_ds_add,
185            exp.TsOrDsToDate: ts_or_ds_to_date_sql("duckdb"),
186            exp.UnixToStr: lambda self, e: f"STRFTIME(TO_TIMESTAMP({self.sql(e, 'this')}), {self.format_time(e)})",
187            exp.UnixToTime: rename_func("TO_TIMESTAMP"),
188            exp.UnixToTimeStr: lambda self, e: f"CAST(TO_TIMESTAMP({self.sql(e, 'this')}) AS TEXT)",
189        }
190
191        TYPE_MAPPING = {
192            **generator.Generator.TYPE_MAPPING,  # type: ignore
193            exp.DataType.Type.BINARY: "BLOB",
194            exp.DataType.Type.CHAR: "TEXT",
195            exp.DataType.Type.FLOAT: "REAL",
196            exp.DataType.Type.NCHAR: "TEXT",
197            exp.DataType.Type.NVARCHAR: "TEXT",
198            exp.DataType.Type.UINT: "UINTEGER",
199            exp.DataType.Type.VARBINARY: "BLOB",
200            exp.DataType.Type.VARCHAR: "TEXT",
201        }
202
203        STAR_MAPPING = {
204            **generator.Generator.STAR_MAPPING,
205            "except": "EXCLUDE",
206        }
207
208        def tablesample_sql(self, expression: exp.TableSample, seed_prefix: str = "SEED") -> str:
209            return super().tablesample_sql(expression, seed_prefix="REPEATABLE")
class DuckDB(sqlglot.dialects.dialect.Dialect):
 79class DuckDB(Dialect):
 80    class Tokenizer(tokens.Tokenizer):
 81        KEYWORDS = {
 82            **tokens.Tokenizer.KEYWORDS,
 83            ":=": TokenType.EQ,
 84            "ATTACH": TokenType.COMMAND,
 85            "BINARY": TokenType.VARBINARY,
 86            "BPCHAR": TokenType.TEXT,
 87            "BITSTRING": TokenType.BIT,
 88            "CHAR": TokenType.TEXT,
 89            "CHARACTER VARYING": TokenType.TEXT,
 90            "EXCLUDE": TokenType.EXCEPT,
 91            "INT1": TokenType.TINYINT,
 92            "LOGICAL": TokenType.BOOLEAN,
 93            "NUMERIC": TokenType.DOUBLE,
 94            "SIGNED": TokenType.INT,
 95            "STRING": TokenType.VARCHAR,
 96            "UBIGINT": TokenType.UBIGINT,
 97            "UINTEGER": TokenType.UINT,
 98            "USMALLINT": TokenType.USMALLINT,
 99            "UTINYINT": TokenType.UTINYINT,
100        }
101
102    class Parser(parser.Parser):
103        FUNCTIONS = {
104            **parser.Parser.FUNCTIONS,  # type: ignore
105            "APPROX_COUNT_DISTINCT": exp.ApproxDistinct.from_arg_list,
106            "ARRAY_LENGTH": exp.ArraySize.from_arg_list,
107            "ARRAY_SORT": exp.SortArray.from_arg_list,
108            "ARRAY_REVERSE_SORT": _sort_array_reverse,
109            "EPOCH": exp.TimeToUnix.from_arg_list,
110            "EPOCH_MS": lambda args: exp.UnixToTime(
111                this=exp.Div(
112                    this=seq_get(args, 0),
113                    expression=exp.Literal.number(1000),
114                )
115            ),
116            "LIST_SORT": exp.SortArray.from_arg_list,
117            "LIST_REVERSE_SORT": _sort_array_reverse,
118            "LIST_VALUE": exp.Array.from_arg_list,
119            "REGEXP_MATCHES": exp.RegexpLike.from_arg_list,
120            "STRFTIME": format_time_lambda(exp.TimeToStr, "duckdb"),
121            "STRPTIME": format_time_lambda(exp.StrToTime, "duckdb"),
122            "STR_SPLIT": exp.Split.from_arg_list,
123            "STRING_SPLIT": exp.Split.from_arg_list,
124            "STRING_TO_ARRAY": exp.Split.from_arg_list,
125            "STR_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
126            "STRING_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
127            "STRUCT_PACK": exp.Struct.from_arg_list,
128            "TO_TIMESTAMP": exp.UnixToTime.from_arg_list,
129            "UNNEST": exp.Explode.from_arg_list,
130        }
131
132        TYPE_TOKENS = {
133            *parser.Parser.TYPE_TOKENS,
134            TokenType.UBIGINT,
135            TokenType.UINT,
136            TokenType.USMALLINT,
137            TokenType.UTINYINT,
138        }
139
140    class Generator(generator.Generator):
141        STRUCT_DELIMITER = ("(", ")")
142
143        TRANSFORMS = {
144            **generator.Generator.TRANSFORMS,  # type: ignore
145            exp.ApproxDistinct: approx_count_distinct_sql,
146            exp.Array: lambda self, e: self.func("ARRAY", e.expressions[0])
147            if isinstance(seq_get(e.expressions, 0), exp.Select)
148            else rename_func("LIST_VALUE")(self, e),
149            exp.ArraySize: rename_func("ARRAY_LENGTH"),
150            exp.ArraySort: _array_sort_sql,
151            exp.ArraySum: rename_func("LIST_SUM"),
152            exp.DataType: _datatype_sql,
153            exp.DateAdd: _date_add,
154            exp.DateDiff: lambda self, e: self.func(
155                "DATE_DIFF", e.args.get("unit") or exp.Literal.string("day"), e.expression, e.this
156            ),
157            exp.DateStrToDate: datestrtodate_sql,
158            exp.DateToDi: lambda self, e: f"CAST(STRFTIME({self.sql(e, 'this')}, {DuckDB.dateint_format}) AS INT)",
159            exp.DiToDate: lambda self, e: f"CAST(STRPTIME(CAST({self.sql(e, 'this')} AS TEXT), {DuckDB.dateint_format}) AS DATE)",
160            exp.Explode: rename_func("UNNEST"),
161            exp.JSONExtract: arrow_json_extract_sql,
162            exp.JSONExtractScalar: arrow_json_extract_scalar_sql,
163            exp.JSONBExtract: arrow_json_extract_sql,
164            exp.JSONBExtractScalar: arrow_json_extract_scalar_sql,
165            exp.LogicalOr: rename_func("BOOL_OR"),
166            exp.Pivot: no_pivot_sql,
167            exp.Properties: no_properties_sql,
168            exp.RegexpExtract: _regexp_extract_sql,
169            exp.RegexpLike: rename_func("REGEXP_MATCHES"),
170            exp.RegexpSplit: rename_func("STR_SPLIT_REGEX"),
171            exp.SafeDivide: no_safe_divide_sql,
172            exp.Split: rename_func("STR_SPLIT"),
173            exp.SortArray: _sort_array_sql,
174            exp.StrPosition: str_position_sql,
175            exp.StrToDate: lambda self, e: f"CAST({str_to_time_sql(self, e)} AS DATE)",
176            exp.StrToTime: str_to_time_sql,
177            exp.StrToUnix: lambda self, e: f"EPOCH(STRPTIME({self.sql(e, 'this')}, {self.format_time(e)}))",
178            exp.Struct: _struct_sql,
179            exp.TimeStrToDate: lambda self, e: f"CAST({self.sql(e, 'this')} AS DATE)",
180            exp.TimeStrToTime: timestrtotime_sql,
181            exp.TimeStrToUnix: lambda self, e: f"EPOCH(CAST({self.sql(e, 'this')} AS TIMESTAMP))",
182            exp.TimeToStr: lambda self, e: f"STRFTIME({self.sql(e, 'this')}, {self.format_time(e)})",
183            exp.TimeToUnix: rename_func("EPOCH"),
184            exp.TsOrDiToDi: lambda self, e: f"CAST(SUBSTR(REPLACE(CAST({self.sql(e, 'this')} AS TEXT), '-', ''), 1, 8) AS INT)",
185            exp.TsOrDsAdd: _ts_or_ds_add,
186            exp.TsOrDsToDate: ts_or_ds_to_date_sql("duckdb"),
187            exp.UnixToStr: lambda self, e: f"STRFTIME(TO_TIMESTAMP({self.sql(e, 'this')}), {self.format_time(e)})",
188            exp.UnixToTime: rename_func("TO_TIMESTAMP"),
189            exp.UnixToTimeStr: lambda self, e: f"CAST(TO_TIMESTAMP({self.sql(e, 'this')}) AS TEXT)",
190        }
191
192        TYPE_MAPPING = {
193            **generator.Generator.TYPE_MAPPING,  # type: ignore
194            exp.DataType.Type.BINARY: "BLOB",
195            exp.DataType.Type.CHAR: "TEXT",
196            exp.DataType.Type.FLOAT: "REAL",
197            exp.DataType.Type.NCHAR: "TEXT",
198            exp.DataType.Type.NVARCHAR: "TEXT",
199            exp.DataType.Type.UINT: "UINTEGER",
200            exp.DataType.Type.VARBINARY: "BLOB",
201            exp.DataType.Type.VARCHAR: "TEXT",
202        }
203
204        STAR_MAPPING = {
205            **generator.Generator.STAR_MAPPING,
206            "except": "EXCLUDE",
207        }
208
209        def tablesample_sql(self, expression: exp.TableSample, seed_prefix: str = "SEED") -> str:
210            return super().tablesample_sql(expression, seed_prefix="REPEATABLE")
class DuckDB.Tokenizer(sqlglot.tokens.Tokenizer):
 80    class Tokenizer(tokens.Tokenizer):
 81        KEYWORDS = {
 82            **tokens.Tokenizer.KEYWORDS,
 83            ":=": TokenType.EQ,
 84            "ATTACH": TokenType.COMMAND,
 85            "BINARY": TokenType.VARBINARY,
 86            "BPCHAR": TokenType.TEXT,
 87            "BITSTRING": TokenType.BIT,
 88            "CHAR": TokenType.TEXT,
 89            "CHARACTER VARYING": TokenType.TEXT,
 90            "EXCLUDE": TokenType.EXCEPT,
 91            "INT1": TokenType.TINYINT,
 92            "LOGICAL": TokenType.BOOLEAN,
 93            "NUMERIC": TokenType.DOUBLE,
 94            "SIGNED": TokenType.INT,
 95            "STRING": TokenType.VARCHAR,
 96            "UBIGINT": TokenType.UBIGINT,
 97            "UINTEGER": TokenType.UINT,
 98            "USMALLINT": TokenType.USMALLINT,
 99            "UTINYINT": TokenType.UTINYINT,
100        }
class DuckDB.Parser(sqlglot.parser.Parser):
102    class Parser(parser.Parser):
103        FUNCTIONS = {
104            **parser.Parser.FUNCTIONS,  # type: ignore
105            "APPROX_COUNT_DISTINCT": exp.ApproxDistinct.from_arg_list,
106            "ARRAY_LENGTH": exp.ArraySize.from_arg_list,
107            "ARRAY_SORT": exp.SortArray.from_arg_list,
108            "ARRAY_REVERSE_SORT": _sort_array_reverse,
109            "EPOCH": exp.TimeToUnix.from_arg_list,
110            "EPOCH_MS": lambda args: exp.UnixToTime(
111                this=exp.Div(
112                    this=seq_get(args, 0),
113                    expression=exp.Literal.number(1000),
114                )
115            ),
116            "LIST_SORT": exp.SortArray.from_arg_list,
117            "LIST_REVERSE_SORT": _sort_array_reverse,
118            "LIST_VALUE": exp.Array.from_arg_list,
119            "REGEXP_MATCHES": exp.RegexpLike.from_arg_list,
120            "STRFTIME": format_time_lambda(exp.TimeToStr, "duckdb"),
121            "STRPTIME": format_time_lambda(exp.StrToTime, "duckdb"),
122            "STR_SPLIT": exp.Split.from_arg_list,
123            "STRING_SPLIT": exp.Split.from_arg_list,
124            "STRING_TO_ARRAY": exp.Split.from_arg_list,
125            "STR_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
126            "STRING_SPLIT_REGEX": exp.RegexpSplit.from_arg_list,
127            "STRUCT_PACK": exp.Struct.from_arg_list,
128            "TO_TIMESTAMP": exp.UnixToTime.from_arg_list,
129            "UNNEST": exp.Explode.from_arg_list,
130        }
131
132        TYPE_TOKENS = {
133            *parser.Parser.TYPE_TOKENS,
134            TokenType.UBIGINT,
135            TokenType.UINT,
136            TokenType.USMALLINT,
137            TokenType.UTINYINT,
138        }

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

Arguments:
  • error_level: the desired error level. Default: ErrorLevel.RAISE
  • error_message_context: determines the amount of context to capture from a query string when displaying the error message (in number of characters). Default: 50.
  • index_offset: Index offset for arrays eg ARRAY[0] vs ARRAY[1] as the head of a list. Default: 0
  • alias_post_tablesample: If the table alias comes after tablesample. Default: False
  • 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
  • null_ordering: Indicates the default null ordering method to use if not explicitly set. Options are "nulls_are_small", "nulls_are_large", "nulls_are_last". Default: "nulls_are_small"
class DuckDB.Generator(sqlglot.generator.Generator):
140    class Generator(generator.Generator):
141        STRUCT_DELIMITER = ("(", ")")
142
143        TRANSFORMS = {
144            **generator.Generator.TRANSFORMS,  # type: ignore
145            exp.ApproxDistinct: approx_count_distinct_sql,
146            exp.Array: lambda self, e: self.func("ARRAY", e.expressions[0])
147            if isinstance(seq_get(e.expressions, 0), exp.Select)
148            else rename_func("LIST_VALUE")(self, e),
149            exp.ArraySize: rename_func("ARRAY_LENGTH"),
150            exp.ArraySort: _array_sort_sql,
151            exp.ArraySum: rename_func("LIST_SUM"),
152            exp.DataType: _datatype_sql,
153            exp.DateAdd: _date_add,
154            exp.DateDiff: lambda self, e: self.func(
155                "DATE_DIFF", e.args.get("unit") or exp.Literal.string("day"), e.expression, e.this
156            ),
157            exp.DateStrToDate: datestrtodate_sql,
158            exp.DateToDi: lambda self, e: f"CAST(STRFTIME({self.sql(e, 'this')}, {DuckDB.dateint_format}) AS INT)",
159            exp.DiToDate: lambda self, e: f"CAST(STRPTIME(CAST({self.sql(e, 'this')} AS TEXT), {DuckDB.dateint_format}) AS DATE)",
160            exp.Explode: rename_func("UNNEST"),
161            exp.JSONExtract: arrow_json_extract_sql,
162            exp.JSONExtractScalar: arrow_json_extract_scalar_sql,
163            exp.JSONBExtract: arrow_json_extract_sql,
164            exp.JSONBExtractScalar: arrow_json_extract_scalar_sql,
165            exp.LogicalOr: rename_func("BOOL_OR"),
166            exp.Pivot: no_pivot_sql,
167            exp.Properties: no_properties_sql,
168            exp.RegexpExtract: _regexp_extract_sql,
169            exp.RegexpLike: rename_func("REGEXP_MATCHES"),
170            exp.RegexpSplit: rename_func("STR_SPLIT_REGEX"),
171            exp.SafeDivide: no_safe_divide_sql,
172            exp.Split: rename_func("STR_SPLIT"),
173            exp.SortArray: _sort_array_sql,
174            exp.StrPosition: str_position_sql,
175            exp.StrToDate: lambda self, e: f"CAST({str_to_time_sql(self, e)} AS DATE)",
176            exp.StrToTime: str_to_time_sql,
177            exp.StrToUnix: lambda self, e: f"EPOCH(STRPTIME({self.sql(e, 'this')}, {self.format_time(e)}))",
178            exp.Struct: _struct_sql,
179            exp.TimeStrToDate: lambda self, e: f"CAST({self.sql(e, 'this')} AS DATE)",
180            exp.TimeStrToTime: timestrtotime_sql,
181            exp.TimeStrToUnix: lambda self, e: f"EPOCH(CAST({self.sql(e, 'this')} AS TIMESTAMP))",
182            exp.TimeToStr: lambda self, e: f"STRFTIME({self.sql(e, 'this')}, {self.format_time(e)})",
183            exp.TimeToUnix: rename_func("EPOCH"),
184            exp.TsOrDiToDi: lambda self, e: f"CAST(SUBSTR(REPLACE(CAST({self.sql(e, 'this')} AS TEXT), '-', ''), 1, 8) AS INT)",
185            exp.TsOrDsAdd: _ts_or_ds_add,
186            exp.TsOrDsToDate: ts_or_ds_to_date_sql("duckdb"),
187            exp.UnixToStr: lambda self, e: f"STRFTIME(TO_TIMESTAMP({self.sql(e, 'this')}), {self.format_time(e)})",
188            exp.UnixToTime: rename_func("TO_TIMESTAMP"),
189            exp.UnixToTimeStr: lambda self, e: f"CAST(TO_TIMESTAMP({self.sql(e, 'this')}) AS TEXT)",
190        }
191
192        TYPE_MAPPING = {
193            **generator.Generator.TYPE_MAPPING,  # type: ignore
194            exp.DataType.Type.BINARY: "BLOB",
195            exp.DataType.Type.CHAR: "TEXT",
196            exp.DataType.Type.FLOAT: "REAL",
197            exp.DataType.Type.NCHAR: "TEXT",
198            exp.DataType.Type.NVARCHAR: "TEXT",
199            exp.DataType.Type.UINT: "UINTEGER",
200            exp.DataType.Type.VARBINARY: "BLOB",
201            exp.DataType.Type.VARCHAR: "TEXT",
202        }
203
204        STAR_MAPPING = {
205            **generator.Generator.STAR_MAPPING,
206            "except": "EXCLUDE",
207        }
208
209        def tablesample_sql(self, expression: exp.TableSample, seed_prefix: str = "SEED") -> str:
210            return super().tablesample_sql(expression, seed_prefix="REPEATABLE")

Generator interprets the given syntax tree and produces a SQL string as an output.

Arguments:
  • time_mapping (dict): the dictionary of custom time mappings in which the key represents a python time format and the output the target time format
  • time_trie (trie): a trie of the time_mapping keys
  • pretty (bool): if set to True the returned string will be formatted. Default: False.
  • quote_start (str): specifies which starting character to use to delimit quotes. Default: '.
  • quote_end (str): specifies which ending character to use to delimit quotes. Default: '.
  • identifier_start (str): specifies which starting character to use to delimit identifiers. Default: ".
  • identifier_end (str): specifies which ending character to use to delimit identifiers. Default: ".
  • identify (bool): if set to True all identifiers will be delimited by the corresponding character.
  • normalize (bool): if set to True all identifiers will lower cased
  • string_escape (str): specifies a string escape character. Default: '.
  • identifier_escape (str): specifies an identifier escape character. Default: ".
  • pad (int): determines padding in a formatted string. Default: 2.
  • indent (int): determines the size of indentation in a formatted string. Default: 4.
  • unnest_column_only (bool): if true unnest table aliases are considered only as column aliases
  • normalize_functions (str): normalize function names, "upper", "lower", or None Default: "upper"
  • alias_post_tablesample (bool): if the table alias comes after tablesample Default: False
  • unsupported_level (ErrorLevel): determines the generator's behavior when it encounters unsupported expressions. Default ErrorLevel.WARN.
  • null_ordering (str): Indicates the default null ordering method to use if not explicitly set. Options are "nulls_are_small", "nulls_are_large", "nulls_are_last". Default: "nulls_are_small"
  • max_unsupported (int): 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 (bool): if the the comma is leading or trailing in select statements 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 or not to preserve comments in the output SQL code. Default: True
def tablesample_sql( self, expression: sqlglot.expressions.TableSample, seed_prefix: str = 'SEED') -> str:
209        def tablesample_sql(self, expression: exp.TableSample, seed_prefix: str = "SEED") -> str:
210            return super().tablesample_sql(expression, seed_prefix="REPEATABLE")
Inherited Members
sqlglot.generator.Generator
Generator
generate
unsupported
sep
seg
pad_comment
maybe_comment
wrap
no_identify
normalize_func
indent
sql
uncache_sql
cache_sql
characterset_sql
column_sql
columndef_sql
columnconstraint_sql
autoincrementcolumnconstraint_sql
compresscolumnconstraint_sql
generatedasidentitycolumnconstraint_sql
notnullcolumnconstraint_sql
primarykeycolumnconstraint_sql
uniquecolumnconstraint_sql
create_sql
describe_sql
prepend_ctes
with_sql
cte_sql
tablealias_sql
bitstring_sql
hexstring_sql
datatype_sql
directory_sql
delete_sql
drop_sql
except_sql
except_op
fetch_sql
filter_sql
hint_sql
index_sql
identifier_sql
national_sql
partition_sql
properties_sql
root_properties
properties
with_properties
locate_properties
property_sql
likeproperty_sql
fallbackproperty_sql
journalproperty_sql
freespaceproperty_sql
afterjournalproperty_sql
checksumproperty_sql
mergeblockratioproperty_sql
datablocksizeproperty_sql
blockcompressionproperty_sql
isolatedloadingproperty_sql
lockingproperty_sql
withdataproperty_sql
insert_sql
intersect_sql
intersect_op
introducer_sql
pseudotype_sql
returning_sql
rowformatdelimitedproperty_sql
table_sql
pivot_sql
tuple_sql
update_sql
values_sql
var_sql
into_sql
from_sql
group_sql
having_sql
join_sql
lambda_sql
lateral_sql
limit_sql
offset_sql
lock_sql
literal_sql
loaddata_sql
null_sql
boolean_sql
order_sql
cluster_sql
distribute_sql
sort_sql
ordered_sql
matchrecognize_sql
query_modifiers
select_sql
schema_sql
star_sql
structkwarg_sql
parameter_sql
sessionparameter_sql
placeholder_sql
subquery_sql
qualify_sql
union_sql
union_op
unnest_sql
where_sql
window_sql
partition_by_sql
window_spec_sql
withingroup_sql
between_sql
bracket_sql
all_sql
any_sql
exists_sql
case_sql
constraint_sql
extract_sql
trim_sql
concat_sql
check_sql
foreignkey_sql
primarykey_sql
unique_sql
if_sql
in_sql
in_unnest_op
interval_sql
return_sql
reference_sql
anonymous_sql
paren_sql
neg_sql
not_sql
alias_sql
aliases_sql
attimezone_sql
add_sql
and_sql
connector_sql
bitwiseand_sql
bitwiseleftshift_sql
bitwisenot_sql
bitwiseor_sql
bitwiserightshift_sql
bitwisexor_sql
cast_sql
currentdate_sql
collate_sql
command_sql
comment_sql
transaction_sql
commit_sql
rollback_sql
altercolumn_sql
renametable_sql
altertable_sql
droppartition_sql
addconstraint_sql
distinct_sql
ignorenulls_sql
respectnulls_sql
intdiv_sql
dpipe_sql
div_sql
floatdiv_sql
overlaps_sql
distance_sql
dot_sql
eq_sql
escape_sql
glob_sql
gt_sql
gte_sql
ilike_sql
is_sql
like_sql
similarto_sql
lt_sql
lte_sql
mod_sql
mul_sql
neq_sql
nullsafeeq_sql
nullsafeneq_sql
or_sql
slice_sql
sub_sql
trycast_sql
use_sql
binary
function_fallback_sql
func
format_args
text_width
format_time
expressions
op_expressions
naked_property
set_operation
tag_sql
token_sql
userdefinedfunction_sql
joinhint_sql
kwarg_sql
when_sql
merge_sql