[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Listener's (very nice) meta-G

From:   Steve Mitchell
To:     Guillaume Cartier    cartier@math.uqam.ca
        David Moon           moon@cambridge.apple.com
cc:     info-mcl
Sub:    Re: Listener's (very nice) meta-G
> From: moon@cambridge.apple.com (David A. Moon)
> > Date: Tue, 21 Apr 92 11:07:48 -0400
> > From: cartier@math.uqam.ca (Guillaume Cartier)
> >
> > Did anyone figure a way to get the dual of meta-G?
> > i.e. you are hitting meta-G furiously, trying to get
> > to a specific expression, but when you finally get to it,
> > your furious hitting has propeled you one or two expressions
> > too far back.
> > Dam!!@#@!#$ you have to start all over again.
> > If only meta-G had a dual key to move you forward...
> Should be meta-minus meta-G, since it takes numeric arguments,
> but trying that signals an error.
Once you've overshot, to get quickly into a position to
start all over again, evaluate something at the prompt,
(type c-g first to get a fresh prompt if necessary), eg
? 1
Now you're back where you started because the position is reset
after each command. (Well almost, because now you have a 1. Anyway,
this saves you having to cycle all the way around with m-g).
If you remember that you hit m-g say around 5 or 6 times before
the entry you want came up, you can type m-5 m-g to get close;
positive numeric arguments _are_ supported (but they won't cause
a wrap-around).
As an alternative, you can use c-m-n and c-m-p to navigate
back and forth to the actual lines you typed into the Listener.
When you get to the line you want, Return will copy it down
(m-Return will copy it down and evaluate it in one step.)
Disclaimer: Availability of these features may depend on the
            version of MCL you have.
Because of the way the input line history is maintained (it's
not a list) it would be hard for us users to write m-- m-g,
but hopefully not hard for the MCL team to accomodate it.
I mention this because the error David Moon reported:
> Error: value nil is not of the expected type number.
> While executing: #<standard-method ccl::ed-grab-last-input (listener)>
could be construed as a bug and if it is somebody should
do something about it anyway.