Unconditional punycode conversion

[Simon Josefsson]

The next-to-last part of IDNA2008-lookup is section 5.5 of RFC 5891:

   The string that has now been validated for lookup is converted to ACE
   form by applying the Punycode algorithm to the string and then adding
   the ACE prefix ("xn--").

Consider an IDNA2008-lookup input label of "foo".  The above appear to
say that this string should be punycode encoded, which seems wrong.

Could someone help me pin-point where it says that section 5.5 does not
apply to an input string "foo"?

Alternatively, should section 5.5 instead say that ACE encoding is only
intended to occur when the label to be converted contains non-ASCII
characters?  That would make more sense.

Thanks,
/Simon