On 9/22/07, david alis <[EMAIL PROTECTED]> wrote:
> Furthermore, because in the 1-item argument case, u isn't invoked
> no protection is possible against such nastiness.

Sure there is.

For example:

   =/&.>'';'a';'ab';'abc'
+-+-+-+-+
|1|a|0|0|
+-+-+-+-+

Let's say I want to change that expression so I get an unboxed
boolean list.  This leaves me with two possible treatments for the
case where the length of the list is 1:

   (1 = =/)&>'';'a';'ab';'abc'
1 0 0 0
   (0 ~: =/)&>'';'a';'ab';'abc'
1 1 0 0

As an aside, note that the result of =/ is going to be 0 for any
character list with more than two items.  Qualitatively
speaking you get four different kinds of answers from =/ on
these kinds of arguments:

length 0 argument -- result of neutral
length 1 argument -- result of identity
length 2 argument -- result of dyad
length 3 argument -- result of non neutral

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

Reply via email to