Review fixes

Author: JATAYU000

openml/_config.py

@@ -495,6 +495,10 @@ def get_cache_directory(self) -> str:
         reversed_url_suffix = os.sep.join(url_parts)  # noqa: PTH118
         return os.path.join(self._root_cache_directory, reversed_url_suffix)  # noqa: PTH118
 
+    def get_minio_download_path(self, url: str) -> str:
+        parsed_url = urlparse(url)
+        return os.path.join(self.get_cache_directory(), "minio", parsed_url.path.lstrip("/"))  #  noqa: PTH118
+
     def set_root_cache_directory(self, root_cache_directory: str | Path) -> None:
         """Set the root cache directory."""
         self._root_cache_directory = Path(root_cache_directory)

openml/datasets/dataset.py

@@ -954,11 +954,11 @@ def _parse_publish_response(self, xml_response: dict) -> None:
         self.dataset_id = int(xml_response["oml:upload_data_set"]["oml:id"])
 
     def publish(self) -> OpenMLDataset:
-        """Publish this flow to OpenML server.
+        """Publish this dataset to the OpenML server.
 
         Returns
         -------
-        self : OpenMLFlow
+        self : OpenMLDataset
         """
         file_elements = self._get_file_elements()
         if "description" not in file_elements:

tests/test_datasets/test_dataset.py

@@ -2,19 +2,14 @@
 from __future__ import annotations
 
 import os
-import unittest.mock
-from pathlib import Path
-import shutil
 from time import time
 
 import numpy as np
 import pandas as pd
 import pytest
-from scipy import sparse
 
 import openml
 from openml.datasets import OpenMLDataFeature, OpenMLDataset
-from openml.exceptions import PyOpenMLError
 from openml.testing import TestBase
 
 import pytest

tests/test_datasets/test_dataset_functions.py

@@ -4,7 +4,6 @@
 import itertools
 import os
 import random
-import shutil
 import time
 import uuid
 from itertools import product
@@ -17,8 +16,6 @@
 import pandas as pd
 import pytest
 import requests
-import requests_mock
-from requests_mock import ANY
 import scipy.sparse
 from oslo_concurrency import lockutils
 
@@ -1870,10 +1867,11 @@ def _dataset_features_is_downloaded(did: int):
 
 def _dataset_data_file_is_downloaded(dataset: OpenMLDataset):
     #TODO to be updated after minio paths is fixed
-    pq_directory = Path(openml.config.get_cache_directory()) / openml.config.get_minio_download_path(dataset._parquet_url)
+    if dataset._parquet_url is not None:
+        pq_directory = Path(openml.config.get_cache_directory()) / Path(openml.config.get_minio_download_path(dataset._parquet_url))
+        if pq_directory.exists():
+            return any(f.suffix == ".pq" for f in pq_directory.iterdir())
     arff_directory = Path(openml.config.get_cache_directory()) / "data/v1/download" / str(dataset.id)
-    if pq_directory.exists():
-        return any(f.suffix == ".pq" for f in pq_directory.iterdir())
     if arff_directory.exists():
         return any(f.suffix == ".arff" for f in arff_directory.iterdir())
     return False
@@ -1894,7 +1892,7 @@ def _assert_datasets_retrieved_successfully(
         - absence of data arff if metadata_only, else it must be present too.
     """
     for dataset in datasets:
-        assert _dataset_description_is_downloaded(dataset)
+        assert _dataset_description_is_downloaded(dataset.id)
 
         has_qualities = _dataset_qualities_is_downloaded(dataset.id)
         assert has_qualities if with_qualities else not has_qualities