Fix stowords filtering

biolab · Sep 30, 2019 · 4916dd9 · 4916dd9
1 parent 70be2b8
commit 4916dd9
Showing 1 changed file with 3 additions and 2 deletions.
diff --git a/orangecontrib/text/preprocess/filter.py b/orangecontrib/text/preprocess/filter.py
@@ -78,7 +78,7 @@ def supported_languages():
         except LookupError:  # when no NLTK data is available
             pass
 
-        return [file.capitalize() for file in stopwords_listdir]
+        return sorted(file.capitalize() for file in stopwords_listdir)
 
     @wait_nltk_data
     def __init__(self, language='English', word_list=None):
@@ -96,7 +96,8 @@ def language(self, value):
         if not self._language:
             self.stopwords = []
         else:
-            self.stopwords = set(stopwords.words(self.language.lower()))
+            self.stopwords = set(
+                x.strip() for x in stopwords.words(self.language.lower()))
 
     def __str__(self):
         config = ''