Although it only works for smallish files, that seems fairly useful for getting as much info as you can out of a corrupt but mostly UTF-8 file?

Any runes that aren't valid will come back as the replacement character. And you can count newlines and print the location of the error(s). You also have the index of the error.