Fix bug with empty part of tags

exa/exa/helpers/utilits.py

@@ -8,12 +8,13 @@ News = namedtuple('News', 'id, tags')
 
 class MLDataMaker:
     def __init__(self, tags_list):
-        self.tags = [self.split_tags(i, ' /') for i in tags_list]
+        self.tags = [self.split_tags(i, '/() ') for i in tags_list]
 
     def split_tags(self, tags, char, replace=' '):
         if len(char) > 1:
             for i in char:
                 tags = tags.replace(i, replace)
+            tags = ' '.join([i for i in tags.split(replace) if len(i) > 0])
             return tags.split(replace)
         return tags.split(char)