small updates to microsoft transformers after PR feedback

pliers/converters/microsoft.py

@@ -10,21 +10,23 @@ class MicrosoftAPITextConverter(MicrosoftVisionAPITransformer, ImageToTextConver
     ''' Detects text within images using the Microsoft Vision API. '''
 
     api_method = 'ocr'
+    _log_attributes = ('api_version', 'language')
+
+    def __init__(self, language='en', **kwargs):
+        self.language = language
+        super(MicrosoftAPITextConverter, self).__init__(**kwargs)
 
     def _convert(self, stim):
         params = {
-            'language': 'en',
+            'language': self.language,
             'detectOrientation': False
         }
         response = self._query_api(stim, params)
 
-        text = ''
+        lines = []
         for r in response['regions']:
             for l in r['lines']:
-                for w in l['words']:
-                    if text:
-                        text += ' ' + w['text']
-                    else:
-                        text += w['text']
+                lines.append(' '.join([w['text'] for w in l['words']]))
 
+        text = '\n'.join(lines)
         return TextStim(text=text, onset=stim.onset, duration=stim.duration)

pliers/tests/converters/test_microsoft_converters.py

@@ -17,4 +17,4 @@ def test_microsoft_vision_api_text_converter():
     conv = MicrosoftAPITextConverter()
     img = ImageStim(join(IMAGE_DIR, 'CC0', '28010844841_c5b81cb9cc_z.jpg'))
     text = conv.transform(img).text
-    assert 'Santander' in text
+    assert 'Santander\nSantander' in text

pliers/transformers/microsoft.py

@@ -16,6 +16,9 @@ class MicrosoftAPITransformer(Transformer):
             Services. Only needs to be passed the first time the extractor is
             initialized.
         location (str): region the subscription key has been registered in.
+            It will be the first part of the endpoint URL suggested by
+            Microsoft when you first created the key.
+            Examples include: westus, westcentralus, eastus
         api_version (str): API version to use.
     '''
 
@@ -49,7 +52,8 @@ def __init__(self, subscription_key=None, location=None,
 
     def _query_api(self, stim, params):
         with stim.get_filename() as filename:
-            data = open(filename, 'rb').read()
+            with open(filename, 'rb') as fp:
+                data = fp.read()
 
         headers = {
             'Content-Type': 'application/octet-stream',
@@ -70,6 +74,9 @@ def _query_api(self, stim, params):
             raise Exception(response['error']['message'])
         elif 'statusCode' in response and response['statusCode'] in [401, 429]:
             raise Exception(response['message'])
+        elif 'code' in response and \
+             response['code'] == 'NotSupportedVisualFeature':
+            raise Exception(response['message'])
 
         return response