Re: [STUMP] Alt-Tab

[Dirk Sondermann]

On 10/25/12 22:19, Ruthard Baudach wrote:
>> == Auszüge aus der Nachricht von  diogofsr vom 2012-10-25 17:52:
>> Ruthard Baudach <address@hidden> writes:
...
>> I didn't try it, but as this is your first time writing Lisp, you might
>> want to make these changes:
>>
>> --- /tmp/alt-tab.lisp    2012-10-25 13:37:35.591757537 -0200
>> +++ /tmp/new.lisp    2012-10-25 13:43:51.118854966 -0200
>> @@ -37,14 +37,14 @@
>>          (setf *alt-tab-index* (mod (1+ *alt-tab-index*) (list-length 
>> *alt-tab-cycle*)))
>>          (select-window-by-number (nth *alt-tab-index* *alt-tab-cycle*)))
>>        ;; else start anew
>> -      (
>> +      (t
>>          (setf *alt-tab-index* 1)
>>          (setf *alt-tab-cycle* (copy-list *alt-tab-list*))
>>      ; if we've fokussed a frame without window, current-window will return 
>> nil => select last used
>>      (cond 
>>        ((not (current-window))
>>         (select-window-by-number (first *alt-tab-cycle*)))
>> -      (
>> +      (t
>>         (push (window-number (current-window)) *alt-tab-cycle*)
>>             (select-window-by-number (second *alt-tab-cycle*))))))
>>
>> The first form in a `cond' is evaluated and if it evaluates to a
>> generalized true, it runs the body.
>>
>> So:
>>
>> (defvar *foo*)
>>
>> (cond ((setf *foo* nil) 4)
>>       (t 2))
>> ; => 2
>>
>> (cond ((setf *foo* 42) 4)
>>       (t 2))
>> ; => 4
> 
> As far as I understood it, cond is the most general if - elsif - else
> construct lisp knows, thus
> 
> (defvar *foo*)
> 
> (cond ((setf *foo* nil) 4)
>       ( 2 ))
> => 2
> 
> At least this works for me using archlinux and the clisp of this
> distribution. But adding the `T' in the last cond - clause might be more
> portable?

Omitting T as the test in the last clause works and is portable,
but, as CLTL expresses it, "is in questionable taste":
http://www.cs.cmu.edu/Groups/AI/html/cltl/clm/node84.html

-- 
Dirk