meerschaum 2.6.0.dev1__py3-none-any.whl → 2.6.2__py3-none-any.whl

meerschaum/utils/dtypes/__init__.py

@@ -259,12 +259,20 @@ def coerce_timezone(
     if isinstance(dt, int):
         return dt
 
-    dt_is_series = hasattr(dt, 'dtype')
+    if isinstance(dt, str):
+        dateutil_parser = mrsm.attempt_import('dateutil.parser')
+        dt = dateutil_parser.parse(dt)
+
+    dt_is_series = hasattr(dt, 'dtype') and hasattr(dt, '__module__')
 
     if dt_is_series:
+        is_dask = 'dask' in dt.__module__
         pandas = mrsm.attempt_import('pandas')
+        dd = mrsm.attempt_import('dask.dataframe') if is_dask else None
         dt_series = (
             pandas.to_datetime(dt, utc=True)
+            if dd is None
+            else dd.to_datetime(dt, utc=True)
         )
         if strip_utc:
             dt_series = dt_series.apply(lambda x: x.replace(tzinfo=None))

meerschaum/utils/dtypes/sql.py

@@ -13,7 +13,7 @@ NUMERIC_PRECISION_FLAVORS: Dict[str, Tuple[int, int]] = {
     'mariadb': (38, 20),
     'mysql': (38, 20),
     'mssql': (28, 10),
-    'duckdb': (15, 4),
+    'duckdb': (15, 3),
     'sqlite': (15, 4),
 }
 TIMEZONE_NAIVE_FLAVORS = {'oracle', 'mysql', 'mariadb'}
@@ -50,6 +50,7 @@ DB_FLAVORS_CAST_DTYPES = {
         'NVARCHAR(2000)': 'NVARCHAR2(2000)',
         'NVARCHAR': 'NVARCHAR2(2000)',
         'NVARCHAR2': 'NVARCHAR2(2000)',
+        'CHAR': 'CHAR(36)',  # UUID columns
     },
     'mssql': {
         'NVARCHAR COLLATE "SQL Latin1 General CP1 CI AS"': 'NVARCHAR(MAX)',
@@ -57,7 +58,6 @@ DB_FLAVORS_CAST_DTYPES = {
         'VARCHAR COLLATE "SQL Latin1 General CP1 CI AS"': 'NVARCHAR(MAX)',
         'VARCHAR COLLATE "SQL_Latin1_General_CP1_CI_AS"': 'NVARCHAR(MAX)',
         'NVARCHAR': 'NVARCHAR(MAX)',
-        'BIT': 'INT',
     },
 }
 for _flavor, (_precision, _scale) in NUMERIC_PRECISION_FLAVORS.items():
@@ -174,7 +174,7 @@ PD_TO_DB_DTYPES_FLAVORS: Dict[str, Dict[str, str]] = {
     'datetime64[ns, UTC]': {
         'timescaledb': 'TIMESTAMPTZ',
         'postgresql': 'TIMESTAMPTZ',
-        'mariadb': 'TIMESTAMP WITH TIME ZONE',
+        'mariadb': 'DATETIME',
         'mysql': 'DATETIME',
         'mssql': 'DATETIMEOFFSET',
         'oracle': 'TIMESTAMP',
@@ -182,7 +182,20 @@ PD_TO_DB_DTYPES_FLAVORS: Dict[str, Dict[str, str]] = {
         'duckdb': 'TIMESTAMPTZ',
         'citus': 'TIMESTAMPTZ',
         'cockroachdb': 'TIMESTAMPTZ',
-        'default': 'TIMESTAMP',
+        'default': 'TIMESTAMPTZ',
+    },
+    'datetime': {
+        'timescaledb': 'TIMESTAMPTZ',
+        'postgresql': 'TIMESTAMPTZ',
+        'mariadb': 'DATETIME',
+        'mysql': 'DATETIME',
+        'mssql': 'DATETIMEOFFSET',
+        'oracle': 'TIMESTAMP',
+        'sqlite': 'TIMESTAMP',
+        'duckdb': 'TIMESTAMPTZ',
+        'citus': 'TIMESTAMPTZ',
+        'cockroachdb': 'TIMESTAMPTZ',
+        'default': 'TIMESTAMPTZ',
     },
     'bool': {
         'timescaledb': 'BOOLEAN',
@@ -256,7 +269,7 @@ PD_TO_DB_DTYPES_FLAVORS: Dict[str, Dict[str, str]] = {
         'mysql': 'CHAR(36)',
         'mssql': 'UNIQUEIDENTIFIER',
         ### I know this is too much space, but erring on the side of caution.
-        'oracle': 'NVARCHAR(2000)',
+        'oracle': 'CHAR(36)',
         'sqlite': 'TEXT',
         'duckdb': 'VARCHAR',
         'citus': 'UUID',
@@ -291,6 +304,19 @@ PD_TO_SQLALCHEMY_DTYPES_FLAVORS: Dict[str, Dict[str, str]] = {
         'cockroachdb': 'Float',
         'default': 'Float',
     },
+    'datetime': {
+        'timescaledb': 'DateTime(timezone=True)',
+        'postgresql': 'DateTime(timezone=True)',
+        'mariadb': 'DateTime(timezone=True)',
+        'mysql': 'DateTime(timezone=True)',
+        'mssql': 'sqlalchemy.dialects.mssql.DATETIMEOFFSET',
+        'oracle': 'sqlalchemy.dialects.oracle.TIMESTAMP(timezone=True)',
+        'sqlite': 'DateTime(timezone=True)',
+        'duckdb': 'DateTime(timezone=True)',
+        'citus': 'DateTime(timezone=True)',
+        'cockroachdb': 'DateTime(timezone=True)',
+        'default': 'DateTime(timezone=True)',
+    },
     'datetime64[ns]': {
         'timescaledb': 'DateTime',
         'postgresql': 'DateTime',
@@ -388,7 +414,7 @@ PD_TO_SQLALCHEMY_DTYPES_FLAVORS: Dict[str, Dict[str, str]] = {
         'mariadb': 'sqlalchemy.dialects.mysql.CHAR(36)',
         'mysql': 'sqlalchemy.dialects.mysql.CHAR(36)',
         'mssql': 'Uuid',
-        'oracle': 'UnicodeText',
+        'oracle': 'sqlalchemy.dialects.oracle.CHAR(36)',
         'sqlite': 'UnicodeText',
         'duckdb': 'UnicodeText',
         'citus': 'Uuid',

meerschaum/utils/misc.py

@@ -975,13 +975,13 @@ def json_serialize_datetime(dt: datetime) -> Union[str, None]:
 
 
 def wget(
-        url: str,
-        dest: Optional[Union[str, 'pathlib.Path']] = None,
-        headers: Optional[Dict[str, Any]] = None,
-        color: bool = True,
-        debug: bool = False,
-        **kw: Any
-    ) -> 'pathlib.Path':
+    url: str,
+    dest: Optional[Union[str, 'pathlib.Path']] = None,
+    headers: Optional[Dict[str, Any]] = None,
+    color: bool = True,
+    debug: bool = False,
+    **kw: Any
+) -> 'pathlib.Path':
     """
     Mimic `wget` with `requests`.
 
@@ -989,7 +989,7 @@ def wget(
     ----------
     url: str
         The URL to the resource to be downloaded.
-        
+
     dest: Optional[Union[str, pathlib.Path]], default None
         The destination path of the downloaded file.
         If `None`, save to the current directory.