bug#5264: 23.1; `repeat' should not hard-code the set of affected commands

[Lars Ingebrigtsen <larsi@gnus.org>]

"Drew Adams" <drew.adams@oracle.com> writes:

> In the definition of command `repeat':
>
> (if (memq last-repeatable-command
>     '(exit-minibuffer
>       minibuffer-complete-and-exit
>       self-insert-and-exit))
>
> This list of affected commands should not be hard-coded.  If user code
> uses different functions from `exit-minibuffer' and
> `minibuffer-complete-and-exit', then users must redefine `repeat' to
> accommodate this command difference.  The `repeat' definition should
> be more flexible than that.
>
> This list of commands should be treated the same way we treat the list
> of dangerous commands (`repeat-too-dangerous'): define a variable
> whose value is the list of commands - do not hard-code the list.

But what it is this does?  If `last-repeatable-command' is any of these
commands, then we...  repeat the first command in `command-history'
instead?  What does that do, exactly?

    (if (memq last-repeatable-command '(exit-minibuffer
					minibuffer-complete-and-exit
					self-insert-and-exit))
        (let ((repeat-command (car command-history)))
          (repeat-message "Repeating %S" repeat-command)
          (eval repeat-command))
      (if (null repeat-arg)
          (repeat-message "Repeating command %S" last-repeatable-command)
        (setq current-prefix-arg repeat-arg)
        (repeat-message
	 "Repeating command %S %S" repeat-arg last-repeatable-command))


-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no