add tagging to flows

Author: amueller

openml/_api_calls.py

@@ -139,7 +139,7 @@ def _parse_server_exception(response):
     additional = None
     if 'oml:additional_information' in server_exception['oml:error']:
         additional = server_exception['oml:error']['oml:additional_information']
-    if code in [370, 372, 512]:
-        # 512 for runs, 370 for datasets (should be 372)
+    if code in [370, 372, 512, 500]:
+        # 512 for runs, 370 for datasets (should be 372), 500 for flows
         return OpenMLServerNoResult(code, message, additional)
     return OpenMLServerException(code, message, additional)

openml/flows/flow.py

@@ -355,6 +355,28 @@ def publish(self):
                              (flow_id, message))
         return self
 
+    def push_tag(self, tag):
+        """Annotates this flow with a tag on the server.
+
+        Parameters
+        ----------
+        tag : str
+            Tag to attach to the flow.
+        """
+        data = {'flow_id': self.flow_id, 'tag': tag}
+        _perform_api_call("/flow/tag", data=data)
+
+    def remove_tag(self, tag):
+        """Removes a tag from this flow on the server.
+
+        Parameters
+        ----------
+        tag : str
+            Tag to attach to the flow.
+        """
+        data = {'flow_id': self.flow_id, 'tag': tag}
+        _perform_api_call("/flow/untag", data=data)
+
 
 def _copy_server_fields(source_flow, target_flow):
     fields_added_by_the_server = ['flow_id', 'uploader', 'version',
@@ -370,5 +392,3 @@ def _copy_server_fields(source_flow, target_flow):
 def _add_if_nonempty(dic, key, value):
     if value is not None:
         dic[key] = value
-
-

openml/flows/functions.py

@@ -4,6 +4,7 @@
 import six
 
 from openml._api_calls import _perform_api_call
+from openml.exceptions import OpenMLServerNoResult
 from . import OpenMLFlow
 
 
@@ -70,7 +71,9 @@ def list_flows(offset=None, size=None, tag=None):
 
 
 def flow_exists(name, external_version):
-    """Retrieves the flow id of the flow uniquely identified by name + external_version.
+    """Retrieves the flow id.
+
+    A flow is uniquely identified by name + external_version.
 
     Parameter
     ---------
@@ -93,8 +96,9 @@ def flow_exists(name, external_version):
     if not (isinstance(name, six.string_types) and len(external_version) > 0):
         raise ValueError('Argument \'version\' should be a non-empty string')
 
-    xml_response = _perform_api_call("flow/exists",
-                                     data={'name': name, 'external_version': external_version})
+    xml_response = _perform_api_call(
+        "flow/exists", data={'name': name, 'external_version':
+                             external_version})
 
     result_dict = xmltodict.parse(xml_response)
     flow_id = int(result_dict['oml:flow_exists']['oml:id'])
@@ -105,15 +109,17 @@ def flow_exists(name, external_version):
 
 
 def _list_flows(api_call):
-    # TODO add proper error handling here!
-    xml_string = _perform_api_call(api_call)
+    try:
+        xml_string = _perform_api_call(api_call)
+    except OpenMLServerNoResult:
+        return []
     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, \
         type(flows_dict['oml:flows'])
     assert flows_dict['oml:flows']['@xmlns:oml'] == \
-           'http://openml.org/openml', flows_dict['oml:flows']['@xmlns:oml']
+        'http://openml.org/openml', flows_dict['oml:flows']['@xmlns:oml']
 
     flows = dict()
     for flow_ in flows_dict['oml:flows']['oml:flow']:
@@ -190,10 +196,10 @@ def assert_flows_equal(flow1, flow2,
         attr2 = getattr(flow2, key, None)
         if key == 'components':
             for name in set(attr1.keys()).union(attr2.keys()):
-                if not name in attr1:
+                if name not in attr1:
                     raise ValueError('Component %s only available in '
                                      'argument2, but not in argument1.' % name)
-                if not name in attr2:
+                if name not in attr2:
                     raise ValueError('Component %s only available in '
                                      'argument2, but not in argument1.' % name)
                 assert_flows_equal(attr1[name], attr2[name],

tests/test_flows/test_flow.py

@@ -3,7 +3,7 @@
 import hashlib
 import re
 import sys
-import time
+from time import time
 
 if sys.version_info[0] >= 3:
     from unittest import mock
@@ -67,6 +67,19 @@ def test_get_flow(self):
         self.assertEqual(subflow_3.parameters['L'], '-1')
         self.assertEqual(len(subflow_3.components), 0)
 
+    def test_tagging(self):
+        flow = openml.flows.get_flow(4024)
+        tag = "testing_tag_{}_{}".format(self.id(), time())
+        flow_list = openml.flows.list_flows(tag=tag)
+        self.assertEqual(len(flow_list), 0)
+        flow.push_tag(tag)
+        flow_list = openml.flows.list_flows(tag=tag)
+        self.assertEqual(len(flow_list), 1)
+        self.assertIn(4024, flow_list)
+        flow.remove_tag(tag)
+        flow_list = openml.flows.list_flows(tag=tag)
+        self.assertEqual(len(flow_list), 0)
+
     def test_from_xml_to_xml(self):
         # Get the raw xml thing
         # TODO maybe get this via get_flow(), which would have to be refactored to allow getting only the xml dictionary