Make Kedro instantiate datasets from kedro_datasetwith higher priority than kedro.extras.datasets

RELEASE.md

@@ -11,8 +11,9 @@
 # Upcoming Release 0.18.4
 
 ## Major features and improvements
-* The config loader objects now implement `UserDict` and the configuration is accessed through `conf_loader['catalog']`
-* You can configure config file patterns through `settings.py` without creating a custom config loader
+* Make Kedro instantiate datasets from `kedro_datasets` with higher priority than `kedro.extras.datasets`. `kedro_datasets` is the namespace for the new `kedro-datasets` python package.
+* The config loader objects now implement `UserDict` and the configuration is accessed through `conf_loader['catalog']`.
+* You can configure config file patterns through `settings.py` without creating a custom config loader.
 
 ## Bug fixes and other changes
 * Fixed `kedro micropkg pull` for packages on PyPI.

kedro/io/core.py

@@ -347,7 +347,7 @@ class Version(namedtuple("Version", ["load", "save"])):
     "intermediate data sets where possible to avoid this warning."
 )
 
-_DEFAULT_PACKAGES = ["kedro.io.", "kedro.extras.datasets.", ""]
+_DEFAULT_PACKAGES = ["kedro.io.", "kedro_datasets.", "kedro.extras.datasets.", ""]
 
 
 def parse_dataset_definition(

tests/io/test_data_catalog.py

@@ -19,7 +19,13 @@
     LambdaDataSet,
     MemoryDataSet,
 )
-from kedro.io.core import VERSION_FORMAT, Version, generate_timestamp
+from kedro.io.core import (
+    _DEFAULT_PACKAGES,
+    VERSION_FORMAT,
+    Version,
+    generate_timestamp,
+    parse_dataset_definition,
+)
 
 
 @pytest.fixture
@@ -373,6 +379,20 @@ def test_config_relative_import(self, sane_config):
         with pytest.raises(DataSetError, match=re.escape(pattern)):
             DataCatalog.from_config(**sane_config)
 
+    def test_config_import_kedro_datasets(self, sane_config, mocker):
+        """Test kedro.extras.datasets default path to the dataset class"""
+        # Spy _load_obj because kedro_datasets is not installed and we can't import it.
+
+        import kedro.io.core  # pylint: disable=import-outside-toplevel
+
+        spy = mocker.spy(kedro.io.core, "_load_obj")
+        parse_dataset_definition(sane_config["catalog"]["boats"])
+        for prefix, call_args in zip(_DEFAULT_PACKAGES, spy.call_args_list):
+            # In Python 3.7 call_args.args is not available thus we access the call
+            # arguments with less meaningful index.
+            # The 1st index returns a tuple, the 2nd index return the name of module.
+            assert call_args[0][0] == f"{prefix}pandas.CSVDataSet"
+
     def test_config_import_extras(self, sane_config):
         """Test kedro.extras.datasets default path to the dataset class"""
         sane_config["catalog"]["boats"]["type"] = "pandas.CSVDataSet"