A Boundary-First Method for Debugging Unicode Escapes

Unicode problems are often diagnosed from the last visible symptom: a log contains \\u65e5, an emoji becomes two replacement characters, or text displays as mojibake. The tempting response is to searc