fix(duckdb)!: transpile timestamp without time zone (#5047)

* fix(duckdb)!: transpile timestamp without time zone

* PR feedback 1 (george)

Author: geooo109

sqlglot/dialects/bigquery.py

@@ -1040,6 +1040,7 @@ class Generator(generator.Generator):
             exp.DataType.Type.SMALLINT: "INT64",
             exp.DataType.Type.TEXT: "STRING",
             exp.DataType.Type.TIMESTAMP: "DATETIME",
+            exp.DataType.Type.TIMESTAMPNTZ: "DATETIME",
             exp.DataType.Type.TIMESTAMPTZ: "TIMESTAMP",
             exp.DataType.Type.TIMESTAMPLTZ: "TIMESTAMP",
             exp.DataType.Type.TINYINT: "INT64",

sqlglot/dialects/clickhouse.py

@@ -1016,6 +1016,7 @@ class Generator(generator.Generator):
             exp.DataType.Type.DECIMAL128: "Decimal128",
             exp.DataType.Type.DECIMAL256: "Decimal256",
             exp.DataType.Type.TIMESTAMP: "DateTime",
+            exp.DataType.Type.TIMESTAMPNTZ: "DateTime",
             exp.DataType.Type.TIMESTAMPTZ: "DateTime",
             exp.DataType.Type.DOUBLE: "Float64",
             exp.DataType.Type.ENUM: "Enum",

sqlglot/dialects/duckdb.py

@@ -321,6 +321,7 @@ class Tokenizer(tokens.Tokenizer):
             "BPCHAR": TokenType.TEXT,
             "CHAR": TokenType.TEXT,
             "CHARACTER VARYING": TokenType.TEXT,
+            "DATETIME": TokenType.TIMESTAMPNTZ,
             "DETACH": TokenType.DETACH,
             "EXCLUDE": TokenType.EXCEPT,
             "LOGICAL": TokenType.BOOLEAN,
@@ -330,6 +331,7 @@ class Tokenizer(tokens.Tokenizer):
             "SIGNED": TokenType.INT,
             "STRING": TokenType.TEXT,
             "SUMMARIZE": TokenType.SUMMARIZE,
+            "TIMESTAMP": TokenType.TIMESTAMPNTZ,
             "TIMESTAMP_S": TokenType.TIMESTAMP_S,
             "TIMESTAMP_MS": TokenType.TIMESTAMP_MS,
             "TIMESTAMP_NS": TokenType.TIMESTAMP_NS,

sqlglot/dialects/hive.py

@@ -503,6 +503,7 @@ class Generator(generator.Generator):
             exp.DataType.Type.ROWVERSION: "BINARY",
             exp.DataType.Type.TEXT: "STRING",
             exp.DataType.Type.TIME: "TIMESTAMP",
+            exp.DataType.Type.TIMESTAMPNTZ: "TIMESTAMP",
             exp.DataType.Type.TIMESTAMPTZ: "TIMESTAMP",
             exp.DataType.Type.UTINYINT: "SMALLINT",
             exp.DataType.Type.VARBINARY: "BINARY",

sqlglot/dialects/mysql.py

@@ -818,6 +818,7 @@ class Generator(generator.Generator):
             exp.DataType.Type.DATETIME2: "DATETIME",
             exp.DataType.Type.SMALLDATETIME: "DATETIME",
             exp.DataType.Type.TIMESTAMP: "DATETIME",
+            exp.DataType.Type.TIMESTAMPNTZ: "DATETIME",
             exp.DataType.Type.TIMESTAMPTZ: "TIMESTAMP",
             exp.DataType.Type.TIMESTAMPLTZ: "TIMESTAMP",
         }

sqlglot/dialects/oracle.py

@@ -266,6 +266,7 @@ class Generator(generator.Generator):
             exp.DataType.Type.NCHAR: "NCHAR",
             exp.DataType.Type.TEXT: "CLOB",
             exp.DataType.Type.TIMETZ: "TIME",
+            exp.DataType.Type.TIMESTAMPNTZ: "TIMESTAMP",
             exp.DataType.Type.TIMESTAMPTZ: "TIMESTAMP",
             exp.DataType.Type.BINARY: "BLOB",
             exp.DataType.Type.VARBINARY: "BLOB",

sqlglot/dialects/postgres.py

@@ -539,6 +539,7 @@ class Generator(generator.Generator):
             exp.DataType.Type.VARBINARY: "BYTEA",
             exp.DataType.Type.ROWVERSION: "BYTEA",
             exp.DataType.Type.DATETIME: "TIMESTAMP",
+            exp.DataType.Type.TIMESTAMPNTZ: "TIMESTAMP",
             exp.DataType.Type.BLOB: "BYTEA",
         }
 

tests/dialects/test_duckdb.py

@@ -428,6 +428,7 @@ def test_duckdb(self):
             "SELECT STRFTIME(CAST('2020-01-01' AS TIMESTAMP), CONCAT('%Y', '%m'))",
             write={
                 "duckdb": "SELECT STRFTIME(CAST('2020-01-01' AS TIMESTAMP), CONCAT('%Y', '%m'))",
+                "spark": "SELECT DATE_FORMAT(CAST('2020-01-01' AS TIMESTAMP_NTZ), CONCAT(COALESCE('yyyy', ''), COALESCE('MM', '')))",
                 "tsql": "SELECT FORMAT(CAST('2020-01-01' AS DATETIME2), CONCAT('yyyy', 'MM'))",
             },
         )
@@ -1110,6 +1111,28 @@ def test_time(self):
             },
         )
 
+        self.validate_all(
+            "SELECT TIMESTAMP 'foo'",
+            write={
+                "duckdb": "SELECT CAST('foo' AS TIMESTAMP)",
+                "hive": "SELECT CAST('foo' AS TIMESTAMP)",
+                "spark2": "SELECT CAST('foo' AS TIMESTAMP)",
+                "spark": "SELECT CAST('foo' AS TIMESTAMP_NTZ)",
+                "postgres": "SELECT CAST('foo' AS TIMESTAMP)",
+                "mysql": "SELECT CAST('foo' AS DATETIME)",
+                "clickhouse": "SELECT CAST('foo' AS Nullable(DateTime))",
+                "databricks": "SELECT CAST('foo' AS TIMESTAMP_NTZ)",
+                "snowflake": "SELECT CAST('foo' AS TIMESTAMPNTZ)",
+                "redshift": "SELECT CAST('foo' AS TIMESTAMP)",
+                "tsql": "SELECT CAST('foo' AS DATETIME2)",
+                "presto": "SELECT CAST('foo' AS TIMESTAMP)",
+                "trino": "SELECT CAST('foo' AS TIMESTAMP)",
+                "oracle": "SELECT CAST('foo' AS TIMESTAMP)",
+                "bigquery": "SELECT CAST('foo' AS DATETIME)",
+                "starrocks": "SELECT CAST('foo' AS DATETIME)",
+            },
+        )
+
     def test_sample(self):
         self.validate_identity(
             "SELECT * FROM tbl USING SAMPLE 5",

tests/dialects/test_mysql.py

@@ -1311,7 +1311,7 @@ def test_safe_div(self):
 
     def test_timestamp_trunc(self):
         hive_dialects = ("spark", "databricks")
-        for dialect in ("postgres", "snowflake", "duckdb", *hive_dialects):
+        for dialect in ("postgres", "snowflake", *hive_dialects):
             for unit in (
                 "SECOND",
                 "DAY",