Add automatic conversion to lists of XML list items in xmltodict call

Author: engelen

openml/datasets/functions.py

@@ -179,7 +179,7 @@ def list_datasets(offset=None, size=None, tag=None):
 def _list_datasets(api_call):
     # TODO add proper error handling here!
     xml_string = _perform_api_call(api_call)
-    datasets_dict = xmltodict.parse(xml_string)
+    datasets_dict = xmltodict.parse(xml_string, force_list=('oml:dataset',))
 
     # Minimalistic check if the XML is useful
     assert type(datasets_dict['oml:data']['oml:dataset']) == list, \
@@ -416,7 +416,7 @@ def _get_dataset_features(did_cache_dir, dataset_id):
         with io.open(features_file, "w", encoding='utf8') as fh:
             fh.write(features_xml)
 
-    features = xmltodict.parse(features_xml)["oml:data_features"]
+    features = xmltodict.parse(features_xml, force_list=('oml:feature',))["oml:data_features"]
 
     return features
 
@@ -452,7 +452,7 @@ def _get_dataset_qualities(did_cache_dir, dataset_id):
         with io.open(qualities_file, "w", encoding='utf8') as fh:
             fh.write(qualities_xml)
 
-    qualities = xmltodict.parse(qualities_xml)['oml:data_qualities']
+    qualities = xmltodict.parse(qualities_xml, force_list=('oml:quality',))['oml:data_qualities']
 
     return qualities
 

openml/flows/functions.py

@@ -107,7 +107,7 @@ def flow_exists(name, external_version):
 def _list_flows(api_call):
     # TODO add proper error handling here!
     xml_string = _perform_api_call(api_call)
-    flows_dict = xmltodict.parse(xml_string)
+    flows_dict = xmltodict.parse(xml_string, force_list=('oml:flow',))
 
     # Minimalistic check if the XML is useful
     assert type(flows_dict['oml:flows']['oml:flow']) == list, \

openml/tasks/functions.py

@@ -128,7 +128,7 @@ def list_tasks(task_type_id=None, offset=None, size=None, tag=None):
 
 def _list_tasks(api_call):
     xml_string = _perform_api_call(api_call)
-    tasks_dict = xmltodict.parse(xml_string)
+    tasks_dict = xmltodict.parse(xml_string, force_list=('oml:task',))
     # Minimalistic check if the XML is useful
     if 'oml:tasks' not in tasks_dict:
         raise ValueError('Error in return XML, does not contain "oml:runs": %s'
@@ -143,6 +143,9 @@ def _list_tasks(api_call):
                          '"http://openml.org/openml": %s'
                          % str(tasks_dict))
 
+    assert type(tasks_dict['oml:tasks']['oml:task']) == list, \
+        type(tasks_dict['oml:tasks'])
+
     tasks = dict()
     procs = _get_estimation_procedure_list()
     proc_dict = dict((x['id'], x) for x in procs)