I've been using Emacs's
iso-accents-mode for a long time to type European languages on ASCII keyboards. Now, for the first time ever I need to type characters that aren't in the "Latin" 8-bit encodings, but available in the Unicode Latin-Extended blocks (in Biblical names such as Măttăthīăs).
It seems much neater to keep using iso-accents for these pseudo-latin characters rather than switch to one of the heavyweight full-unicode input methods. Is there a language definition (or could I create one) that would turn e.g. _i into ī and -a into ă, or is
iso-accents-mode fundamentally restricted to 8-bit chars?