Python's encode () and decode () methods accept an 'errors' parameter where you can specify strategies like 'ignore', 'replace', or 'xmlcharrefreplace' to handle these issues gracefully without ...

Fix for issue python-mode#211: 'Non-ascii characters in docstrings trigger unicode error'. Use _get_encoding to encode docs in show_doc ...

This is fixed in python3 branch. The Python DLL is now loaded using the wide-string Windows API, so any unicode path is allowed.