> If function-key-map is nil but key-translation-map is not then further
> down uninitialised values of keybuf are scanned (fkey.start = 31):
>
> Does this patch fix it?
I don't understand the details but it seems to DTRT, so yes. A few
trivial things:
>...
> ! /* This is a submap reached by looking up, in PARENT,
> ! the events from START to END. */
> ! Lisp_Object map;
> ! /* Positions [START, END) in the key sequence buffer
[START, END] ?
> ! are the key that we have scanned so far.
> ! Those events are the ones that we will replace
> ! if PAREHT maps them into a key sequence. */
PARENT
> int start, end;
> } keyremap;
>...
> delayed_switch_frame = Qnil;
> fkey.map = fkey.parent = Vfunction_key_map;
> keytran.map = keytran.parent = Vkey_translation_map;
> ! fkey.start = fkey.end = 0;
Add comment /* If there is no key-translation-map, turn off scanning. */ ?
> keytran.start = keytran.end = KEYMAPP (keytran.map) ? 0 : bufsize + 1;
>
> if (INTERACTIVE)
--
Nick http://www.inet.net.nz/~nickrob
_______________________________________________
emacs-pretest-bug mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
