Huh. Not
obvious to me what he is doing in bookloupe.
Bookloupe was intended to be the UTF-8-aware replacement for
gutcheck. However, if I remember correctly, at the time the author
wrote it PG wanted it to report all the same errors as gutcheck
would or they would not agree to switch over to using it.
So it ended up only partially UTF-8-aware, as with full UTF-8
awareness it wouldn't report some errors that gutcheck would have
reported. But as far as I know Bookloupe is what PG now uses, not
gutcheck.