Need clear behavior for strings with utf8 flag set about bencode HOT 1 OPEN

The UTF8 flag is irrelevant, and whether the string was bytes or characters cannot be known outside of one particular case. Namely, if the string matches /[^\x0-\xff]/, then it contains some wide characters, so it cannot be a (proper) byte string. (It may contain bytes mixed in with the characters if the code constructing it is buggy.) But a string that does not match this could be anything, regardless of whether its UTF8 flag is set.

In any case you are right, the docs should declare how the module handles this issue.

from bencode.