ENH O(1) stop-word lookup when list provided

The docstring says stop_words can be a list, but it should be accessed as a set.

Author: jnothman

sklearn/feature_extraction/tests/test_text.py

@@ -280,7 +280,7 @@ def test_countvectorizer_stop_words():
     assert_raises(ValueError, cv.get_stop_words)
     stoplist = ['some', 'other', 'words']
     cv.set_params(stop_words=stoplist)
-    assert_equal(cv.get_stop_words(), stoplist)
+    assert_equal(cv.get_stop_words(), set(stoplist))
 
 
 def test_countvectorizer_empty_vocabulary():

sklearn/feature_extraction/text.py

@@ -88,8 +88,10 @@ def _check_stop_list(stop):
         return ENGLISH_STOP_WORDS
     elif isinstance(stop, six.string_types):
         raise ValueError("not a built-in stop list: %s" % stop)
+    elif stop is None:
+        return None
     else:               # assume it's a collection
-        return stop
+        return frozenset(stop)
 
 
 class VectorizerMixin(object):