Also you had a left argument of 7 in your first example but not in
your later examples.

(Which wouldn't change the error report but would change the process
which lead to that error.)

Thanks,

-- 
Raul
On Tue, Sep 18, 2018 at 6:32 PM 'Pascal Jasmin' via Programming
<[email protected]> wrote:
>
> oops sorry, my misunderstanding of ^:v1`v2.
>
> v1 operates on y directly.  not the result of v2.
>
>
>
> ________________________________
> From: 'Pascal Jasmin' via Programming <[email protected]>
> To: Programming Forum <[email protected]>
> Sent: Tuesday, September 18, 2018 6:24 PM
> Subject: [Jprogramming] bug with ^:v1`v2
>
>
>
> this seems like a bug at least. tested with 8.05, 8.06 and 8.07.
>
>
> > 7 ]^:('' -: 0 {:: ]) (1 {. boxopen@]) 2 NB. fine
>
> 2
>
>
>
> ]^:(('' -: 0 {:: ])`(1 {. boxopen@])) 2 NB. should be identical to above
>
> |length error
>
>
>
> > ]^:(('' -: 0 {:: ])`(1 {. boxopen@])) ,2
>
> 2
>
>
>
> the core issue is:
>
>
> 0 {:: 2
>
> |length error
>
> |   0    {::2
>
> 0 {:: 1 {. 2
>
> 2
>
>
>
> the bug though is that v2 should be called before v1 sees its result, and v1 
> may be reporting the length error as if it were seeing the raw argument?
>
> ----------------------------------------------------------------------
>
> For information about J forums see http://www.jsoftware.com/forums.htm
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to