Fix completion for String config type.

Since 2a705e2eb6 non-specialized config types are
String. However, String had an overloaded complete() which defaulted to
returning None.

Now we use the normal complete() which relies on valid_values if completions
isn't given instead.

Fixes #1223.

CHANGELOG.asciidoc

@@ -23,6 +23,11 @@ Added
 - New `--quiet` argument for the `:debug-pyeval` command to not open a tab with
   the results. Note `:debug-pyeval` is still only intended for debugging.
 
+Fixed
+~~~~~
+
+- Fixed completion for various config values when using `:set`.
+
 v0.5.0
 ------
 

qutebrowser/config/configtypes.py

@@ -270,7 +270,10 @@ class String(BaseType):
                                             "long!".format(self.maxlen))
 
     def complete(self):
-        return self._completions
+        if self._completions is not None:
+            return self._completions
+        else:
+            return super().complete()
 
 
 class List(BaseType):

tests/unit/config/test_configtypes.py

@@ -335,6 +335,15 @@ class TestString:
     def test_complete(self, klass, value):
         assert klass(completions=value).complete() == value
 
+    @pytest.mark.parametrize('valid_values, expected', [
+        (configtypes.ValidValues('one', 'two'),
+            [('one', ''), ('two', '')]),
+        (configtypes.ValidValues(('1', 'one'), ('2', 'two')),
+            [('1', 'one'), ('2', 'two')]),
+    ])
+    def test_complete_valid_values(self, klass, valid_values, expected):
+        assert klass(valid_values=valid_values).complete() == expected
+
 
 class TestList: