[ENH] Add get_cache_size Utility Function

openml/utils.py

@@ -427,6 +427,18 @@ def safe_func(*args: P.args, **kwargs: P.kwargs) -> R:
         return func
 
 
+def get_cache_size() -> int:
+    """Calculate the size of OpenML cache directory
+
+    Returns
+    -------
+    cache_size: int
+        Total size of cache in bytes
+    """
+    path = Path(config.get_cache_directory())
+    return sum(f.stat().st_size for f in path.rglob("*") if f.is_file())
+
+
 def _create_lockfiles_dir() -> Path:
     path = Path(config.get_cache_directory()) / "locks"
     # TODO(eddiebergman): Not sure why this is allowed to error and ignore???

tests/test_utils/test_utils.py

@@ -152,3 +152,28 @@ def test_correct_test_server_download_state():
     task = openml.tasks.get_task(119)
     dataset = task.get_dataset()
     assert len(dataset.features) == dataset.get_data()[0].shape[1]
+
+@unittest.mock.patch("openml.config.get_cache_directory")
+def test_get_cache_size(config_mock,tmp_path):
+    """
+    Test that the OpenML cache size utility correctly reports the cache directory
+    size before and after fetching a dataset.
+
+    This test uses a temporary directory (tmp_path) as the cache location by
+    patching the configuration via config_mock. It verifies two conditions:
+    empty cache and after dataset fetch. 
+
+    Parameters
+    ----------
+    config_mock : unittest.mock.Mock
+         A mock that overrides the configured cache directory to point to tmp_path.
+    tmp_path : pathlib.Path
+         A pytest-provided temporary directory used as an isolated cache location.
+    """
+
+    config_mock.return_value = tmp_path
+    cache_size = openml.utils.get_cache_size()
+    assert cache_size == 0
+    dataset = openml.datasets.get_dataset(dataset_id=3)
+    cache_size = openml.utils.get_cache_size()
+    assert cache_size == 2009