--- Raul Miller wrote:
> On 8/6/07, Sherlock, Ric wrote:
> > I'm not that keen on all the Caps in there. Can it be cleaned 
> > up/simplified more?
> 
> A fork of the form
>    [: u v
> 
> is equivalent to
>    u @: v
> 
> Note however, that if v is a composite verb you usually need 
> to include it in parenthesis.
<useful pointers deleted>
> Hopefully, you can take it from there...

Yes thanks, I had tried substituting @: for [: but had got stuck on the
brackets around v.
I should have looked back at your earlier reply to a question of mine
about [: & @: 
http://www.jsoftware.com/pipermail/programming/2007-July/007322.html

So I get the following that are equivalent
direxist1=: 'd' e."1 [: > [: , [: ({:"1)
1!:0@(fboxname&>)@(dropPS&.>)@boxopen
direxist2=: 'd' e."1
>@:,@:({:"1@:(1!:0@(fboxname&>)@:(dropPS&.>)@:boxopen))

After all that I have to say I find the [: version easier to parse, and
given that the only real difference is style (from earlier thread), I
think I'm leaning towards the [: version.

I suppose, in my original question, I wasn't really intending to
restrict answers to using the operations/algorithm I had come up with.
Are there other ideas of how to tacitly test for directory existence?
 
> Note also that if you have linear display on, you can use 
> that to remind you where parenthesis are unnecessary

Yes I currently like to have linear & boxed on together.
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to