feat: create column types for polars data types (#208)

Closes partially #196.

### Summary of Changes

* Add `polars`
* Create `ColumnType` for `polars` data type
* Create `Schema` for `polars` data frame

---------

Co-authored-by: megalinter-bot <[email protected]>

pyproject.toml

@@ -20,6 +20,7 @@ pandas = "^2.0.0"
 pillow = "^9.5.0"
 scikit-learn = "^1.2.0"
 seaborn = "^0.12.2"
+polars = {extras = ["pandas", "pyarrow", "xlsx2csv"], version = "^0.17.5"}
 
 [tool.poetry.group.dev.dependencies]
 pytest = "^7.2.1"

src/safeds/data/tabular/containers/_column.py

@@ -46,7 +46,7 @@ def __init__(self, name: str, data: Iterable, type_: ColumnType | None = None) -
         self._name: str = name
         self._data: pd.Series = data if isinstance(data, pd.Series) else pd.Series(data)
         # noinspection PyProtectedMember
-        self._type: ColumnType = type_ if type_ is not None else ColumnType._from_numpy_dtype(self._data.dtype)
+        self._type: ColumnType = type_ if type_ is not None else ColumnType._from_numpy_data_type(self._data.dtype)
 
     def __eq__(self, other: object) -> bool:
         if not isinstance(other, Column):

src/safeds/data/tabular/containers/_row.py

@@ -47,7 +47,7 @@ def from_dict(data: dict[str, Any]) -> Row:
         """
         row_frame = pd.DataFrame([data.values()], columns=list(data.keys()))
         # noinspection PyProtectedMember
-        return Row(data.values(), Schema._from_dataframe(row_frame))
+        return Row(data.values(), Schema._from_pandas_dataframe(row_frame))
 
     # ------------------------------------------------------------------------------------------------------------------
     # Dunder methods
@@ -65,7 +65,7 @@ def __init__(self, data: Iterable, schema: Schema | None = None):
             dataframe = self._data.to_frame().T
             dataframe.columns = column_names
             # noinspection PyProtectedMember
-            self._schema = Schema._from_dataframe(dataframe)
+            self._schema = Schema._from_pandas_dataframe(dataframe)
 
     def __eq__(self, other: Any) -> bool:
         if not isinstance(other, Row):
@@ -136,7 +136,7 @@ def get_value(self, column_name: str) -> Any:
         if not self._schema.has_column(column_name):
             raise UnknownColumnNameError([column_name])
         # noinspection PyProtectedMember
-        return self._data[self._schema._get_column_index_by_name(column_name)]
+        return self._data[self._schema._get_column_index(column_name)]
 
     def has_column(self, column_name: str) -> bool:
         """

src/safeds/data/tabular/containers/_table.py

@@ -233,7 +233,7 @@ def __init__(self, data: Iterable, schema: Schema | None = None):
         | [from_rows][safeds.data.tabular.containers._table.Table.from_rows]           | Create a table from a list of rows.    |
         """
         self._data: pd.DataFrame = data if isinstance(data, pd.DataFrame) else pd.DataFrame(data)
-        self._schema: Schema = Schema._from_dataframe(self._data) if schema is None else schema
+        self._schema: Schema = Schema._from_pandas_dataframe(self._data) if schema is None else schema
 
         if self._data.empty:
             self._data = pd.DataFrame(columns=self._schema.get_column_names())
@@ -305,7 +305,7 @@ def get_column(self, column_name: str) -> Column:
         if self._schema.has_column(column_name):
             output_column = Column(
                 column_name,
-                self._data.iloc[:, [self._schema._get_column_index_by_name(column_name)]].squeeze(),
+                self._data.iloc[:, [self._schema._get_column_index(column_name)]].squeeze(),
                 self._schema.get_type_of_column(column_name),
             )
             return output_column

src/safeds/data/tabular/typing/_column_type.py

@@ -4,44 +4,63 @@
 from dataclasses import dataclass
 from typing import TYPE_CHECKING
 
+from polars import FLOAT_DTYPES as POLARS_FLOAT_DTYPES
+from polars import INTEGER_DTYPES as POLARS_INTEGER_DTYPES
+from polars import TEMPORAL_DTYPES as POLARS_TEMPORAL_DTYPES
+from polars import Boolean as PolarsBoolean
+from polars import Decimal as PolarsDecimal
+from polars import Object as PolarsObject
+from polars import PolarsDataType
+from polars import Utf8 as PolarsUtf8
+
 if TYPE_CHECKING:
     import numpy as np
 
 
 class ColumnType(ABC):
     """Abstract base class for column types."""
 
-    @abstractmethod
-    def is_nullable(self) -> bool:
+    @staticmethod
+    def _from_numpy_data_type(data_type: np.dtype) -> ColumnType:
         """
-        Return whether the given column type is nullable.
+        Return the column type for a given `numpy` data type.
+
+        Parameters
+        ----------
+        data_type : numpy.dtype
+            The `numpy` data type.
 
         Returns
         -------
-        is_nullable : bool
-            True if the column is nullable.
-        """
+        column_type : ColumnType
+            The ColumnType.
 
-    @abstractmethod
-    def is_numeric(self) -> bool:
+        Raises
+        ------
+        NotImplementedError
+            If the given data type is not supported.
         """
-        Return whether the given column type is numeric.
+        if data_type.kind in ("u", "i"):
+            return Integer()
+        if data_type.kind == "b":
+            return Boolean()
+        if data_type.kind == "f":
+            return RealNumber()
+        if data_type.kind in ("S", "U", "O", "M", "m"):
+            return String()
 
-        Returns
-        -------
-        is_numeric : bool
-            True if the column is numeric.
-        """
+        message = f"Unsupported numpy data type '{data_type}'."
+        raise NotImplementedError(message)
 
     @staticmethod
-    def _from_numpy_dtype(dtype: np.dtype) -> ColumnType:
+    def _from_polars_data_type(data_type: PolarsDataType) -> ColumnType:
         """
-        Return the column type for a given numpy dtype.
+        Return the column type for a given `polars` data type.
 
         Parameters
         ----------
-        dtype : numpy.dtype
-            The numpy dtype.
+        data_type : PolarsDataType
+            The `polars` data type.
 
         Returns
         -------
@@ -50,18 +69,42 @@ def _from_numpy_dtype(dtype: np.dtype) -> ColumnType:
 
         Raises
         ------
-        TypeError
-            If the given dtype is not supported.
+        NotImplementedError
+            If the given data type is not supported.
         """
-        if dtype.kind in ("u", "i"):
+        if data_type in POLARS_INTEGER_DTYPES:
             return Integer()
-        if dtype.kind == "b":
+        if data_type is PolarsBoolean:
             return Boolean()
-        if dtype.kind == "f":
+        if data_type in POLARS_FLOAT_DTYPES or data_type is PolarsDecimal:
             return RealNumber()
-        if dtype.kind in ("S", "U", "O", "M", "m"):
+        if data_type is PolarsUtf8 or data_type is PolarsObject or data_type in POLARS_TEMPORAL_DTYPES:
             return String()
-        raise TypeError("Unexpected column type")
+
+        message = f"Unsupported polars data type '{data_type}'."
+        raise NotImplementedError(message)
+
+    @abstractmethod
+    def is_nullable(self) -> bool:
+        """
+        Return whether the given column type is nullable.
+
+        Returns
+        -------
+        is_nullable : bool
+            True if the column is nullable.
+        """
+
+    @abstractmethod
+    def is_numeric(self) -> bool:
+        """
+        Return whether the given column type is numeric.
+
+        Returns
+        -------
+        is_numeric : bool
+            True if the column is numeric.
+        """
 
 
 @dataclass