> Wow, that was fast. I received your reply 10 minutes
> after I sent my question, so your implementation and
> testing had to take considerably less time than that.
FYI: the changes to the source to implement the function
are as follows:
...
switch(v->id){
case CPOUND: f1=jtiota1; break;
...
The C function that implements f\ already had a big switch
for various special cases. The new case CPOUND
adds the hook for the #\ case. f1 is the C function
that implements the monad for f\ .
The interpreter already had a function apv(n,a,b)
for computing the arithmetic progression vector a+b*i.n
and, for an array w, the macro IC(w) (item count)
gives #w as a "long". Therefore:
static DF1(jtiota1){R apv(IC(w),1L,1L);}
Similar thing for #\., with
static DF1(jtiota1rev){R apv(IC(w),IC(w),-1L);}
I did not bother to add anything to the test scripts
for these changes.
----- Original Message -----
From: Raul Miller <[email protected]>
Date: Thursday, August 5, 2010 9:24
Subject: Re: [Jchat] #\
To: Chat forum <[email protected]>
> On Thu, Aug 5, 2010 at 11:56 AM, Roger Hui <[email protected]> wrote:
> > Done.
> >
> > #\ 10 4$'ten four'
> > 1 2 3 4 5 6 7 8 9 10
> >
> > x=: i.1e6
> > ts '#\x'
> > 0.00933387 4.19501e6
>
> Wow, that was fast. I received your reply 10 minutes
> after I sent my question, so your implementation and
> testing had to take considerably less time than that.
>
> Thank you!
>
> That said, if you have not already done so, perhaps you
> could at some point give #\. similar treatment?
>
> (I believe I have used >:@i...@-@# occasionally, though
> I do not have a convenient way of searching for references.)
>
> Thanks again,
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
