Change the default of use_exact_op_types in the tokenizer.

parso/python/diff.py

@@ -308,7 +308,7 @@ class DiffParser(object):
         is_first_token = True
         omitted_first_indent = False
         indents = []
-        tokens = generate_tokens(lines, use_exact_op_types=True)
+        tokens = generate_tokens(lines)
         stack = self._active_parser.pgen_parser.stack
         for typ, string, start_pos, prefix in tokens:
             start_pos = start_pos[0] + line_offset, start_pos[1]