Merge pull request #148 from openml/fix/uploads

FIX upload all xml files as files (not as post argument)

Author: amueller

openml/_api_calls.py

@@ -7,7 +7,7 @@
 from .exceptions import OpenMLServerError
 
 
-def _perform_api_call(call, data=None, file_dictionary=None,
+def _perform_api_call(call, file_dictionary=None,
                       file_elements=None, add_authentication=True):
     """
     Perform an API call at the OpenML server.
@@ -18,11 +18,14 @@ def _read_url(self, url, add_authentication=False, data=None, filePath=None):
     ----------
     call : str
         The API call. For example data/list
-    data : dict (default=None)
-        Dictionary containing data which will be sent to the OpenML
-        server via a POST request.
-    **kwargs
-        Further arguments which are appended as GET arguments.
+    file_dictionary : dict
+        Mapping of {filename: path} of files which should be uploaded to the
+        server.
+    file_elements : dict
+        Mapping of {filename: str} of strings which should be uploaded as
+        files to the server.
+    add_authentication : bool
+        Whether to add authentication (api key) to the request.
 
     Returns
     -------
@@ -36,16 +39,16 @@ def _read_url(self, url, add_authentication=False, data=None, filePath=None):
         url += "/"
     url += call
     if file_dictionary is not None or file_elements is not None:
-        return _read_url_files(url, data=data, file_dictionary=file_dictionary,
+        return _read_url_files(url, file_dictionary=file_dictionary,
                                file_elements=file_elements)
-    return _read_url(url, data=data)
+    return _read_url(url)
 
 
-def _read_url_files(url, data=None, file_dictionary=None, file_elements=None):
+def _read_url_files(url, file_dictionary=None, file_elements=None):
     """do a post request to url with data None, file content of
     file_dictionary and sending file_elements as files"""
-    if data is None:
-        data = {}
+
+    data = {}
     data['api_key'] = config.apikey
     if file_elements is None:
         file_elements = {}
@@ -78,9 +81,9 @@ def _read_url_files(url, data=None, file_dictionary=None, file_elements=None):
     return response.status_code, response.text
 
 
-def _read_url(url, data=None):
-    if data is None:
-        data = {}
+def _read_url(url):
+
+    data = {}
     data['api_key'] = config.apikey
 
     # Using requests.post sets header 'Accept-encoding' automatically to

openml/datasets/dataset.py

@@ -1,7 +1,8 @@
 import gzip
+import logging
 import os
 import sys
-import logging
+
 import arff
 
 import numpy as np
@@ -267,12 +268,17 @@ def publish(self):
         return_value : string
             xml return from server
         """
-        data = {'description': self._to_xml()}
+
+        file_elements = {'description': self._to_xml()}
+        file_dictionary = {}
+
         if self.data_file is not None:
-            return_code, return_value = _perform_api_call(
-                "/data/", data=data, file_dictionary={'dataset': self.data_file})
-        else:
-            return_code, return_value = _perform_api_call("/data/", data=data)
+            file_dictionary['dataset'] = self.data_file
+
+        return_code, return_value = _perform_api_call(
+            "/data/", file_dictionary=file_dictionary,
+            file_elements=file_elements)
+
         return return_code, return_value
 
     def _to_xml(self):

openml/flows/flow.py

@@ -81,9 +81,10 @@ def publish(self):
         https://github.com/openml/website/blob/master/openml_OS/views/pages/rest_api/xsd/openml.implementation.upload.xsd
         """
         xml_description = self._generate_flow_xml()
-        data = {'description': xml_description, 'source': self.source}
+
+        file_elements = {'description': xml_description} #, 'source': self.source}
         return_code, return_value = _perform_api_call(
-            "/flow/", data=data)
+            "flow/", file_elements=file_elements)
         return return_code, return_value
 
     def _ensure_flow_exists(self):
@@ -137,7 +138,7 @@ def _check_flow_exists(name, version):
         raise ValueError('Argument \'version\' should be a non-empty string')
 
     return_code, xml_response = _perform_api_call(
-        "/flow/exists/%s/%s" % (name, version))
+        "flow/exists/%s/%s" % (name, version))
     # TODO check with latest version of code if this raises an exception
     if return_code != 200:
         # fixme raise appropriate error

openml/runs/run.py

@@ -79,10 +79,10 @@ def publish(self):
         """
         predictions = arff.dumps(self._generate_arff())
         description_xml = self._create_description_xml()
-        data = {'predictions': ("predictions.csv", predictions),
-                'description': ("description.xml", description_xml)}
+        file_elements = {'predictions': ("predictions.csv", predictions),
+                         'description': ("description.xml", description_xml)}
         return_code, return_value = _perform_api_call(
-            "/run/", file_elements=data)
+            "/run/", file_elements=file_elements)
         return return_code, return_value
 
     def _create_description_xml(self):

tests/datasets/test_datasets.py

@@ -98,6 +98,7 @@ def test_list_datasets_by_tag(self):
             self.assertIn(dataset['status'], ['in_preparation', 'active',
                                               'deactivated'])
 
+    @unittest.skip('See https://github.com/openml/openml-python/issues/149')
     def test_check_datasets_active(self):
         active = openml.datasets.check_datasets_active([1, 17])
         self.assertTrue(active[1])

tests/flows/test_flow.py

@@ -17,6 +17,7 @@ def check_flow(flow):
         for flow in flows:
             check_flow(flow)
 
+    @unittest.skip('Not tested until test sentinels are added back.')
     def test_upload_flow(self):
         flow = openml.OpenMLFlow(model=DummyClassifier(), description="test description")
         return_code, return_value = flow.publish()