Provide clearer error when server provides bad data description XML (#1178)

Author: PGijsbers

openml/_api_calls.py

@@ -23,6 +23,14 @@
 )
 
 
+def _create_url_from_endpoint(endpoint: str) -> str:
+    url = config.server
+    if not url.endswith("/"):
+        url += "/"
+    url += endpoint
+    return url.replace("=", "%3d")
+
+
 def _perform_api_call(call, request_method, data=None, file_elements=None):
     """
     Perform an API call at the OpenML server.
@@ -50,12 +58,7 @@ def _perform_api_call(call, request_method, data=None, file_elements=None):
     return_value : str
         Return value of the OpenML server
     """
-    url = config.server
-    if not url.endswith("/"):
-        url += "/"
-    url += call
-
-    url = url.replace("=", "%3d")
+    url = _create_url_from_endpoint(call)
     logging.info("Starting [%s] request for the URL %s", request_method, url)
     start = time.time()
 

openml/datasets/functions.py

@@ -3,6 +3,7 @@
 import io
 import logging
 import os
+from pyexpat import ExpatError
 from typing import List, Dict, Union, Optional, cast
 
 import numpy as np
@@ -19,6 +20,7 @@
 from .dataset import OpenMLDataset
 from ..exceptions import (
     OpenMLHashException,
+    OpenMLServerError,
     OpenMLServerException,
     OpenMLPrivateDatasetError,
 )
@@ -437,7 +439,7 @@ def get_dataset(
             parquet_file = None
         remove_dataset_cache = False
     except OpenMLServerException as e:
-        # if there was an exception,
+        # if there was an exception
         # check if the user had access to the dataset
         if e.code == 112:
             raise OpenMLPrivateDatasetError(e.message) from None
@@ -949,14 +951,18 @@ def _get_dataset_description(did_cache_dir, dataset_id):
     try:
         with io.open(description_file, encoding="utf8") as fh:
             dataset_xml = fh.read()
+        description = xmltodict.parse(dataset_xml)["oml:data_set_description"]
     except Exception:
         url_extension = "data/{}".format(dataset_id)
         dataset_xml = openml._api_calls._perform_api_call(url_extension, "get")
+        try:
+            description = xmltodict.parse(dataset_xml)["oml:data_set_description"]
+        except ExpatError as e:
+            url = openml._api_calls._create_url_from_endpoint(url_extension)
+            raise OpenMLServerError(f"Dataset description XML at '{url}' is malformed.") from e
         with io.open(description_file, "w", encoding="utf8") as fh:
             fh.write(dataset_xml)
 
-    description = xmltodict.parse(dataset_xml)["oml:data_set_description"]
-
     return description
 
 

tests/test_datasets/test_dataset_functions.py

@@ -1240,7 +1240,7 @@ def _wait_for_dataset_being_processed(self, dataset_id):
             try:
                 downloaded_dataset = openml.datasets.get_dataset(dataset_id)
                 break
-            except Exception as e:
+            except OpenMLServerException as e:
                 # returned code 273: Dataset not processed yet
                 # returned code 362: No qualities found
                 TestBase.logger.error(