Re: Reading char choices

[Chong Yidong]

"Davis Herring" <address@hidden> writes:

> Couldn't we pass the prompt to read-event?  Also, how does this
> inhibit quitting out of the whole function?
> ...
> If I'm missing something and we really do inhibit quitting, surely
> this should be (unless inhibit-keyboard-quit ...).

Fixed, thanks.  Seems that dired-query, from which I factored out this
code, was a bit buggy.

>> ((setq done (memq char chars))))))
>
> So if it's not valid, we just ignore it.  Would we rather display a
> message (as for `y-or-n-p'), at least optionally?

That's fine if someone wants to implement it.  Currently, we don't do
this for the risky local variables prompt, nor for dired-query.