PandasDataFrame IO

packages/ordeq-pandas/src/ordeq_pandas/__init__.py

@@ -1,5 +1,6 @@
 from ordeq_pandas.csv import PandasCSV
+from ordeq_pandas.dataframe import PandasDataFrame
 from ordeq_pandas.excel import PandasExcel
 from ordeq_pandas.parquet import PandasParquet
 
-__all__ = ("PandasCSV", "PandasExcel", "PandasParquet")
+__all__ = ("PandasCSV", "PandasDataFrame", "PandasExcel", "PandasParquet")

packages/ordeq-pandas/src/ordeq_pandas/dataframe.py

@@ -0,0 +1,56 @@
+from collections.abc import Iterable
+from dataclasses import dataclass, field
+from typing import Any
+from uuid import uuid4
+
+import pandas as pd
+from ordeq import Input
+from pandas import Index
+
+
+@dataclass(frozen=True, kw_only=True)
+class PandasDataFrame(Input[pd.DataFrame]):
+    """Allows a Pandas DataFrame to be hard-coded as IO. This is suitable
+    for small tables such as simple dimension tables that are unlikely to
+    change. It is also useful for unit testing.
+
+    Example usage:
+
+    ```pycon
+    >>> from ordeq_pandas import PandasDataFrame
+    >>> df = PandasDataFrame(
+    ...     data=(
+    ...         (2022, "file_2022.xlsx"),
+    ...         (2023, "file_2023.xlsx"),
+    ...         (2024, "file_2024.xlsx"),
+    ...     ),
+    ...     columns=("year", "datafile"),
+    ... ).load()
+    >>> print(df.shape)
+    (3, 2)
+
+    ```
+
+    """
+
+    _idx: str = field(
+        init=False, default_factory=lambda: str(uuid4()), repr=False
+    )
+    data: Iterable = field(hash=False)
+    columns: Index | Iterable[str] | None = field(hash=False, default=None)
+
+    def load(self, **load_options: Any) -> pd.DataFrame:
+        """Loads the DataFrame from the provided data and columns.
+
+        Args:
+            **load_options: Additional options passed to `pd.DataFrame`.
+
+        Returns:
+            Loaded Pandas DataFrame.
+        """
+
+        return pd.DataFrame(
+            data=self.data,
+            columns=self.columns,  # type: ignore[arg-type]
+            **load_options,
+        )

packages/ordeq-pandas/tests/test_dataframe.py

@@ -0,0 +1,37 @@
+import pandas as pd
+from ordeq_pandas import PandasDataFrame
+
+
+def test_it_loads():
+    data = [
+        (2022, "file_2022.xlsx"),
+        (2023, "file_2023.xlsx"),
+        (2024, "file_2024.xlsx"),
+    ]
+    columns = ["year", "datafile"]
+    expected = pd.DataFrame(data, columns=columns)
+    actual = PandasDataFrame(data=data, columns=columns).load()
+    pd.testing.assert_frame_equal(actual, expected)
+
+
+def test_its_unique():
+    data = [
+        (2022, "file_2022.xlsx"),
+        (2023, "file_2023.xlsx"),
+        (2024, "file_2024.xlsx"),
+    ]
+    columns = ["year", "datafile"]
+    df1 = PandasDataFrame(data=data, columns=columns)
+    df2 = PandasDataFrame(data=data, columns=columns)
+    assert df1 != df2
+
+
+def test_its_hashable():
+    data = [
+        (2022, "file_2022.xlsx"),
+        (2023, "file_2023.xlsx"),
+        (2024, "file_2024.xlsx"),
+    ]
+    columns = ["year", "datafile"]
+    df = PandasDataFrame(data=data, columns=columns)
+    assert isinstance(hash(df), int)