update the extract_comments.py script to match xml_converter. will be a temporary fix, the two scripts are easily mergeable