reorganize keyword creation for legacy tokenizer

- only save partial words without internal spaces
- consider comma and semicolon a separator of full words
- consider parts before an opening bracket a full word
  (but not the part after the bracket)

Fixes #244.

test/python/test_tokenizer_legacy_icu.py

@@ -223,11 +223,32 @@ def test_update_special_phrase_modify(analyzer, word_table):
 
 
 def test_process_place_names(analyzer, getorcreate_term_id):
-
     with analyzer() as anl:
         info = anl.process_place({'name' : {'name' : 'Soft bAr', 'ref': '34'}})
 
-    assert info['names'] == '{1,2,3,4,5,6}'
+    assert info['names'] == '{1,2,3,4,5}'
+
+
+@pytest.mark.parametrize('sep', [',' , ';'])
+def test_full_names_with_separator(analyzer, getorcreate_term_id, sep):
+    with analyzer() as anl:
+        full_names =
+            anl._compute_full_names({'name' : sep.join(('New York', 'Big Apple'))})
+
+        expect = set((anl.make_standard_word(w) for w in ('New York', 'Big Apple')))
+
+    assert full_names == expect
+
+
+def test_process_place_names_with_bracket(analyzer, getorcreate_term_id):
+    with analyzer() as anl:
+        info = anl.process_place({'name' :
+                                   {'name' : 'Houseboat (left)'}})
+
+        expect = set((anl.make_standard_word(w) for w in
+                       (' houseboat', ' houseboat left', 'houseboat', 'left')))
+
+    assert eval(info['names']) == expect
 
 
 @pytest.mark.parametrize('pcode', ['12345', 'AB 123', '34-345'])