This question is a follow-up to this answer (q.v.).
I understand that Emacs interprets esc as
M- only when the keystroke that follows it corresponds to a "character" (e.g. A, 5, ;, etc.), as opposed to, e.g., ←, F5, home, etc.
What can I add to my
.emacs file to extend the esc ⇒
M- interpretation to the case where the following key is a "non-character" key?
OK, the following snippet takes care of my immediate problem, but it's not an answer to my question (see below):
(add-hook 'org-mode-hook (lambda () (define-key org-mode-map [27 S-down] 'org-shiftmetadown) (define-key org-mode-map [27 S-up] 'org-shiftmetaup) (define-key org-mode-map [27 S-right] 'org-shiftmetaright) (define-key org-mode-map [27 S-left] 'org-shiftmetaleft) (define-key org-mode-map [27 down] 'org-metadown) (define-key org-mode-map [27 up] 'org-metaup) (define-key org-mode-map [27 right] 'org-metaright) (define-key org-mode-map [27 left] 'org-metaleft) (define-key org-mode-map [27 return] 'org-meta-return) (define-key org-mode-map [27 S-return] 'org-insert-todo-heading)))
To find the bindings I needed to define, I got a listing of
C-h v), and searched for
M- to get all the bindings beginning with
(M-S-down . org-shiftmetadown) (M-S-up . org-shiftmetaup) (M-S-right . org-shiftmetaright) (M-S-left . org-shiftmetaleft) (M-down . org-metadown) (M-up . org-metaup) (M-right . org-metaright) (M-left . org-metaleft) (M-return . org-meta-return) (M-S-return . org-insert-todo-heading)
Each one of these lines is the basis for a corresponding
define-key expression in the mode-hook shown above. The transformation should be pretty obvious. (The 27 in the
define-key expressions is the code for esc.)
One would need to do this sort of thing for every single mode X that defines a key binding of the form
M- followed by a non-character key sequence. Specifically, for every such binding, one would have to include a
define-key expression in a mode-hook for mode X. Therefore, this approach is obviously a very laborious, fragile, and hard-to-maintain hack. It does not do what the question requests, but instead painstakingly simulates it for certain special cases. For this reason I did not post it as an answer, in the hopes that someone will post an actual answer to the original question.