Re: called-interactively-p and edebug

[Le Wang <l26wang@gmail.com>]

[-- Attachment #1: Type: text/plain, Size: 1793 bytes --]

I should add for posterity's sake that interactive forms do work in
defadvice:

(defun foo (beg end)
  (interactive "r")
  (message "%s %s" beg end))

(defadvice foo (around foo-adv activate)
  (interactive
   (list (point-min) (point-max)))
  ad-do-it)

On Wed, Feb 2, 2011 at 9:14 AM, Le Wang <l26wang@gmail.com> wrote:

> Thanks Stefan, I discovered this in a defadvice, but reported it using the
> simplest repro I could find.  I'll keep your tip about preferring
> interactive forms in mind.
>
>
> On Wed, Feb 2, 2011 at 1:00 AM, Stefan Monnier <monnier@iro.umontreal.ca>wrote:
>
>> > When I step through this function:
>>
>> > (defun foo ()
>> >   (interactive)
>> >   (when (called-interactively-p 'any)
>> >     (message "Interactive!")
>> >     'foo-called-interactively))
>>
>> > (called-interactively-p 'any) is nil even when called interactively.  Is
>> > this a bug?
>>
>> Yes, it's a known bug: stepping through a function with Edebug is done
>> by rewriting the function, and called-interactively-p is a hackish
>> function that doesn't have a clean enough semantics to survive
>> this rewrite.
>> That's one of the reasons the docstring says:
>>
>>   This function is meant for implementing advice and other
>>   function-modifying features.  Instead of using this, it is sometimes
>>   cleaner to give your function an extra optional argument whose
>>   `interactive' spec specifies non-nil unconditionally ("p" is a good
>>   way to do this), or via (not (or executing-kbd-macro noninteractive)).
>>
>> So I'd recommend you use the interactive spec instead, as in:
>>
>>  (defun foo (am-i-interactive)
>>    (interactive "p")
>>    (when am-i-interactive
>>       (message "Interactive!")
>>      'foo-called-interactively))
>>
>>
>> -- Stefan
>>
>
>
>
> --
> Le
>



-- 
Le

[-- Attachment #2: Type: text/html, Size: 2793 bytes --]