OP is in fact about doing your best to detect encoding without proper tagging. It is not perfect and you won't have 100% accuracy. But it can get pretty decent

OP is in fact about doing your best to detect encoding without proper tagging. It is not perfect and you won't have 100% accuracy. But it can get pretty decent (OP is literally about the techniques used to do so, statistically etc), and is necessary because of the actual world we live in. If just refusing to do it would get MS to export in UTF-8 by default that would of course have happened a long time ago!