Originally UTF-8, then decoded as Latin-1, then re-encoded to UTF-8. You can recompute the exact bytes in the page with this Python code: >>> 'Zürich'.enco

Originally UTF-8, then decoded as Latin-1, then re-encoded to UTF-8. You can recompute the exact bytes in the page with this Python code: >>> 'Zürich'.encode('utf-8').decode('latin1').encode('utf-8') b'Z\xc3\x83\xc2\xbcrich'