On Fri, Oct 7, 2011 at 10:31 AM, Linda Alvord <[email protected]> wrote:
> Your solution assumes that you knew what the author was thinking. I guess
> you have to solve the total problem yourself and hope you get a simpler
> solution.

Yes, if I am going to be creating meaningful names this requires that
I understand the relevant meanings.

But we can create explicit definitions without meaningful names -- the
only thing we need to know, for that, is the basic grammar involved
(in other words: is the definition used as a monadic verb or a dyadic
verb or both?).

That said, if I was building a program to convert a tacit verb to an
explicit verb, I would not be using the trace facility.  That would be
too much work for me.  Instead, I would start from the atomic
representation of the tacit verb.  And I would be giving special
treatment to boxed sequences beginning with a box that matches
e.&(<;._1'/2/3'), working from the outside in.

And the hardest part of this problem would be name management.  And
even that part would not be too hard, since I can use arbitrary names.

-- 
Raul
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to