fix: fix mypy issue with ExcelDataNode._read_as_dataframe() may return a dict and add tests for custom exposed type to be a list of pandas and numpy

taipy/core/data/excel.py

@@ -153,7 +153,7 @@ class ExcelDataNode(DataNode, _FileDataNodeMixin, _TabularDataNodeMixin):
         return self._read_from_path()
 
     def _read_from_path(self, path: Optional[str] = None, **read_kwargs) -> Any:
-        if not path:
+        if path is None:
             path = self._path
 
         if self.properties[self._EXPOSED_TYPE_PROPERTY] == self._EXPOSED_TYPE_PANDAS:
@@ -198,7 +198,7 @@ class ExcelDataNode(DataNode, _FileDataNodeMixin, _TabularDataNodeMixin):
 
                     if isinstance(sheet_exposed_type, str):
                         if sheet_exposed_type == self._EXPOSED_TYPE_NUMPY:
-                            work_books[sheet_name] = self._read_as_pandas_dataframe(path, sheet_name).to_numpy()
+                            work_books[sheet_name] = self._read_as_numpy(path, sheet_name)
                         elif sheet_exposed_type == self._EXPOSED_TYPE_PANDAS:
                             work_books[sheet_name] = self._read_as_pandas_dataframe(path, sheet_name)
                         continue
@@ -220,8 +220,8 @@ class ExcelDataNode(DataNode, _FileDataNodeMixin, _TabularDataNodeMixin):
 
         return work_books
 
-    def _read_as_numpy(self, path: str):
-        sheets = self._read_as_pandas_dataframe(path=path)
+    def _read_as_numpy(self, path: str, sheet_names=None):
+        sheets = self._read_as_pandas_dataframe(path=path, sheet_names=sheet_names)
         if isinstance(sheets, dict):
             return {sheet_name: df.to_numpy() for sheet_name, df in sheets.items()}
         return sheets.to_numpy()

tests/core/data/test_read_excel_data_node.py

@@ -58,6 +58,7 @@ class MyCustomObject2:
 excel_file_path = os.path.join(pathlib.Path(__file__).parent.resolve(), "data_sample/example.xlsx")
 sheet_names = ["Sheet1", "Sheet2"]
 custom_class_dict = {"Sheet1": MyCustomObject1, "Sheet2": MyCustomObject2}
+custom_pandas_numpy_exposed_type_dict = {"Sheet1": "pandas", "Sheet2": "numpy"}
 
 
 def test_raise_no_data_with_header():
@@ -400,7 +401,7 @@ def test_read_multi_sheet_with_header_single_custom_exposed_type():
             assert row_custom_no_sheet_name.text == row_custom.text
 
 
-def test_read_multi_sheet_with_header_multiple_custom_exposed_type():
+def test_read_multi_sheet_with_header_multiple_custom_object_exposed_type():
     data_pandas = pd.read_excel(excel_file_path, sheet_name=sheet_names)
 
     # With sheet name
@@ -461,6 +462,48 @@ def test_read_multi_sheet_with_header_multiple_custom_exposed_type():
             assert row_custom_no_sheet_name.text == row_custom.text
 
 
+def test_read_multi_sheet_with_header_multiple_custom_pandas_numpy_exposed_type():
+    # With sheet name
+    excel_dn_as_pandas_numpy = ExcelDataNode(
+        "bar",
+        Scope.SCENARIO,
+        properties={
+            "path": excel_file_path,
+            "sheet_name": sheet_names,
+            "exposed_type": custom_pandas_numpy_exposed_type_dict,
+        },
+    )
+    assert excel_dn_as_pandas_numpy.properties["exposed_type"] == custom_pandas_numpy_exposed_type_dict
+    multi_data_custom = excel_dn_as_pandas_numpy.read()
+    assert isinstance(multi_data_custom["Sheet1"], pd.DataFrame)
+    assert isinstance(multi_data_custom["Sheet2"], np.ndarray)
+
+    excel_dn_as_pandas_numpy = ExcelDataNode(
+        "bar",
+        Scope.SCENARIO,
+        properties={
+            "path": excel_file_path,
+            "sheet_name": sheet_names,
+            "exposed_type": ["pandas", "numpy"],
+        },
+    )
+    assert excel_dn_as_pandas_numpy.properties["exposed_type"] == ["pandas", "numpy"]
+    multi_data_custom = excel_dn_as_pandas_numpy.read()
+    assert isinstance(multi_data_custom["Sheet1"], pd.DataFrame)
+    assert isinstance(multi_data_custom["Sheet2"], np.ndarray)
+
+    # Without sheet name
+    excel_dn_as_pandas_numpy_no_sheet_name = ExcelDataNode(
+        "bar",
+        Scope.SCENARIO,
+        properties={"path": excel_file_path, "exposed_type": custom_pandas_numpy_exposed_type_dict},
+    )
+    assert excel_dn_as_pandas_numpy_no_sheet_name.properties["exposed_type"] == custom_pandas_numpy_exposed_type_dict
+    multi_data_custom_no_sheet_name = excel_dn_as_pandas_numpy_no_sheet_name.read()
+    assert isinstance(multi_data_custom_no_sheet_name["Sheet1"], pd.DataFrame)
+    assert isinstance(multi_data_custom_no_sheet_name["Sheet2"], np.ndarray)
+
+
 def test_read_multi_sheet_without_header_pandas():
     # With sheet name
     excel_data_node_as_pandas = ExcelDataNode(
@@ -525,7 +568,7 @@ def test_read_multi_sheet_without_header_numpy():
         assert np.array_equal(data_numpy[key], data_numpy_no_sheet_name[key])
 
 
-def test_read_multi_sheet_without_header_single_custom_exposed_type():
+def test_read_multi_sheet_without_header_single_custom_object_exposed_type():
     data_pandas = pd.read_excel(excel_file_path, header=None, sheet_name=sheet_names)
 
     # With sheet name
@@ -579,7 +622,7 @@ def test_read_multi_sheet_without_header_single_custom_exposed_type():
             assert row_custom_no_sheet_name.text == row_custom.text
 
 
-def test_read_multi_sheet_without_header_multiple_custom_exposed_type():
+def test_read_multi_sheet_without_header_multiple_custom_object_exposed_type():
     data_pandas = pd.read_excel(excel_file_path, header=None, sheet_name=sheet_names)
 
     # With sheet names
@@ -643,3 +686,51 @@ def test_read_multi_sheet_without_header_multiple_custom_exposed_type():
             assert row_custom_no_sheet_name.id == row_custom.id
             assert row_custom_no_sheet_name.integer == row_custom.integer
             assert row_custom_no_sheet_name.text == row_custom.text
+
+
+def test_read_multi_sheet_without_header_multiple_custom_pandas_numpy_exposed_type():
+    # With sheet names
+    excel_dn_as_pandas_numpy = ExcelDataNode(
+        "bar",
+        Scope.SCENARIO,
+        properties={
+            "path": excel_file_path,
+            "sheet_name": sheet_names,
+            "exposed_type": custom_pandas_numpy_exposed_type_dict,
+            "has_header": False,
+        },
+    )
+    assert excel_dn_as_pandas_numpy.properties["exposed_type"] == custom_pandas_numpy_exposed_type_dict
+    multi_data_custom = excel_dn_as_pandas_numpy.read()
+    assert isinstance(multi_data_custom["Sheet1"], pd.DataFrame)
+    assert isinstance(multi_data_custom["Sheet2"], np.ndarray)
+
+    excel_dn_as_pandas_numpy = ExcelDataNode(
+        "bar",
+        Scope.SCENARIO,
+        properties={
+            "path": excel_file_path,
+            "sheet_name": sheet_names,
+            "exposed_type": ["pandas", "numpy"],
+            "has_header": False,
+        },
+    )
+    assert excel_dn_as_pandas_numpy.properties["exposed_type"] == ["pandas", "numpy"]
+    multi_data_custom = excel_dn_as_pandas_numpy.read()
+    assert isinstance(multi_data_custom["Sheet1"], pd.DataFrame)
+    assert isinstance(multi_data_custom["Sheet2"], np.ndarray)
+
+    # Without sheet names
+    excel_dn_as_pandas_numpy_no_sheet_name = ExcelDataNode(
+        "bar",
+        Scope.SCENARIO,
+        properties={
+            "path": excel_file_path,
+            "has_header": False,
+            "exposed_type": custom_pandas_numpy_exposed_type_dict,
+        },
+    )
+    multi_data_custom_no_sheet_name = excel_dn_as_pandas_numpy_no_sheet_name.read()
+    multi_data_custom_no_sheet_name = excel_dn_as_pandas_numpy.read()
+    assert isinstance(multi_data_custom_no_sheet_name["Sheet1"], pd.DataFrame)
+    assert isinstance(multi_data_custom_no_sheet_name["Sheet2"], np.ndarray)