Key completion for dataset access

RELEASE.md

@@ -6,6 +6,8 @@
 * Updated error message for invalid catalog entries.
 * Updated error message for catalog entries when the dataset class is not found with hints on how to resolve the issue.
 * Fixed a bug in the `DataCatalog` `shallow_copy()` method to ensure it returns the type of the used catalog and doesn't cast it to `DataCatalog`.
+* Implemented key completion support for accessing datasets in the `DataCatalog`.
+
 
 ## Breaking changes to the API
 

kedro/io/data_catalog.py

@@ -105,28 +105,36 @@ def __init__(
         """Return a _FrozenDatasets instance from some datasets collections.
         Each collection could either be another _FrozenDatasets or a dictionary.
         """
+        self._original_names: set[str] = set()
         for collection in datasets_collections:
             if isinstance(collection, _FrozenDatasets):
                 self.__dict__.update(collection.__dict__)
+                self._original_names.update(collection._original_names)
             else:
                 # Non-word characters in dataset names are replaced with `__`
                 # for easy access to transcoded/prefixed datasets.
-                self.__dict__.update(
-                    {
-                        _sub_nonword_chars(dataset_name): dataset
-                        for dataset_name, dataset in collection.items()
-                    }
-                )
+                for dataset_name, dataset in collection.items():
+                    self.__dict__[_sub_nonword_chars(dataset_name)] = dataset
+                    self._original_names.add(dataset_name)
 
     # Don't allow users to add/change attributes on the fly
     def __setattr__(self, key: str, value: Any) -> None:
+        if key == "_original_names":
+            super().__setattr__(key, value)
+            return
         msg = "Operation not allowed! "
         if key in self.__dict__:
             msg += "Please change datasets through configuration."
         else:
             msg += "Please use DataCatalog.add() instead."
         raise AttributeError(msg)
 
+    def _ipython_key_completions_(self) -> list[str]:
+        return list(self._original_names)
+
+    def __getitem__(self, key: str) -> Any:
+        return self.__dict__[_sub_nonword_chars(key)]
+
 
 class DataCatalog:
     """``DataCatalog`` stores instances of ``AbstractDataset`` implementations

tests/io/test_data_catalog.py

@@ -442,6 +442,30 @@ class MyDataCatalog(DataCatalog):
         copy = data_catalog.shallow_copy()
         assert isinstance(copy, MyDataCatalog)
 
+    def test_key_completions(self, data_catalog_from_config):
+        """Test catalog.datasets key completions"""
+        assert isinstance(data_catalog_from_config.datasets["boats"], CSVDataset)
+        assert isinstance(data_catalog_from_config.datasets["cars"], CSVDataset)
+        data_catalog_from_config.add_feed_dict(
+            {
+                "params:model_options": [1, 2, 4],
+                "params:model_options.random_state": [0, 42, 67],
+            }
+        )
+        assert isinstance(
+            data_catalog_from_config.datasets["params:model_options"], MemoryDataset
+        )
+        assert isinstance(
+            data_catalog_from_config.datasets["params__model_options.random_state"],
+            MemoryDataset,
+        )
+        assert set(data_catalog_from_config.datasets._ipython_key_completions_()) == {
+            "boats",
+            "cars",
+            "params:model_options",
+            "params:model_options.random_state",
+        }
+
 
 class TestDataCatalogFromConfig:
     def test_from_sane_config(self, data_catalog_from_config, dummy_dataframe):