rgom

I will do that, of course. For the time being feel free to revert the change. I can see 3 test cases failing. At least two of them used the...

Ok, here it comes: 1. My initial fix was incorrect - sorry for that. 2. I haven't initially tested (executed) whole org.python.pydev.editor.codecompletion.PythonCompletionWithoutBuiltinsTest 3. My newly added test was incorrect -...

Oh, that's fine. The patch tried to address just that case and could be missing some global picture. If the risk of introducing it is too high, please disregard it....