build(datasets): add type targets for return types (#797)

* build(datasets): circumvent Requests ref not found See also: sphinx-doc/sphinx#9653 Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): circumvent dask_expr ref not found Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): add VersionNotFoundError exc type Signed-off-by: Deepyaman Datta <[email protected]> * docs(datasets): update malformed dataset docstring Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): add missing DataFrame type target Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): add missing DataFrame type target Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): add missing DataFrame type target Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): add missing transformers.Pipeline Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): create new `ir.Table` type target Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): add a `numpy.ndarray` type target Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): add type target for `plt.Figure`s Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): create NetworkX Graph type target Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): add missing DataFrame type target Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): add `PIL.Image.Image` type target Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): add type target for Plotly Figure Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): add type target for Figure Widget Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): add missing DataFrame type target Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): add missing LazyFrame type target Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): add missing DataFrame type target Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): add missing target for DeltaTable Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): add missing DataFrame type target Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): create type target for csr_matrix Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): add type target for Keras `Model` Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): add missing type target for video Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): make `type_targets` for Anthropic Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): add type target for Cohere models Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): add xarray dataset `type_targets` Signed-off-by: Deepyaman Datta <[email protected]> * build(datasets): add target for xarray `DataArray` Signed-off-by: Deepyaman Datta <[email protected]> * feat(datasets)!: expose `load` and `save` publicly Signed-off-by: Deepyaman Datta <[email protected]> * fix(datasets): replace calls to _meth with wrapped Signed-off-by: Deepyaman Datta <[email protected]> * fix(datasets): pass `self` for `__wrapped__` calls Signed-off-by: Deepyaman Datta <[email protected]> * revert(datasets): restore to `_load`s and `_save`s Refs: 3204803, 42201cf, b40597c Signed-off-by: Deepyaman Datta <[email protected]> --------- Signed-off-by: Deepyaman Datta <[email protected]>
kedro-org · Aug 2, 2024 · 95219ad · 95219ad
1 parent 0c35e78
commit 95219ad
Show file tree

Hide file tree

Showing 2 changed files with 36 additions and 5 deletions.
diff --git a/kedro-datasets/docs/source/conf.py b/kedro-datasets/docs/source/conf.py
@@ -107,20 +107,49 @@
         "AbstractDataset",
         "kedro.io.core.Version",
         "requests.auth.AuthBase",
+        "requests.models.Response",
         "google.oauth2.credentials.Credentials",
         "deltalake.table.Metadata",
         "DataCatalog",
         "ibis.backends.BaseBackend",
         "langchain_openai.chat_models.base.ChatOpenAI",
         "langchain_openai.embeddings.base.OpenAIEmbeddings",
+        "dask_expr._collection.DataFrame",
+        "DataFrame",
+        "pd.DataFrame",
+        "geopandas.geodataframe.GeoDataFrame",
+        "transformers.pipelines.base.Pipeline",
+        "ibis.expr.types.relations.Table",
+        "numpy.ndarray",
+        "matplotlib.figure.Figure",
+        "networkx.classes.graph.Graph",
+        "pandas.core.frame.DataFrame",
+        "PIL.Image.Image",
+        "plotly.graph_objs._figure.Figure",
+        "plotly.graph_objs._figurewidget.FigureWidget",
+        "polars.dataframe.frame.DataFrame",
+        "polars.lazyframe.frame.LazyFrame",
+        "snowflake.snowpark.dataframe.DataFrame",
+        "delta.tables.DeltaTable",
+        "pyspark.sql.dataframe.DataFrame",
+        "scipy.sparse._csr.csr_matrix",
+        "keras.src.models.model.Model",
+        "kedro_datasets.video.video_dataset.AbstractVideo",
+        "langchain_anthropic.chat_models.ChatAnthropic",
+        "langchain_cohere.chat_models.ChatCohere",
+        "xarray.core.dataset.Dataset",
+        "xarray.core.dataarray.DataArray",
     ),
     "py:data": (
         "typing.Any",
         "typing.Union",
         "typing.Optional",
         "typing.Tuple",
     ),
-    "py:exc": ("DatasetError",),
+    "py:exc": (
+        "DatasetError",
+        "VersionNotFoundError",
+    ),
 }
 # https://stackoverflow.com/questions/61770698/sphinx-nit-picky-mode-but-only-for-links-i-explicitly-wrote
 nitpick_ignore = [(key, value) for key in type_targets for value in type_targets[key]]

diff --git a/kedro-datasets/kedro_datasets/databricks/managed_table_dataset.py b/kedro-datasets/kedro_datasets/databricks/managed_table_dataset.py
@@ -45,10 +45,12 @@ class ManagedTable:
 
     def __post_init__(self):
         """Run validation methods if declared.
+
         The validation method can be a simple check
         that raises DatasetError.
-        The validation is performed by calling a function named:
-            `validate_<field_name>(self, value) -> raises DatasetError`
+
+        The validation is performed by calling a function with the signature
+        `validate_<field_name>(self, value) -> raises DatasetError`.
         """
         for name in self.__dataclass_fields__.keys():
             method = getattr(self, f"_validate_{name}", None)
@@ -293,11 +295,11 @@ def _load(self) -> DataFrame | pd.DataFrame:
 
         Raises:
             VersionNotFoundError: if the version defined in
-            the init doesn't exist
+                the init doesn't exist
 
         Returns:
             Union[DataFrame, pd.DataFrame]: Returns a dataframe
-            in the format defined in the init
+                in the format defined in the init
         """
         if self._version and self._version.load >= 0:
             try: