Raul wrote:
> You can use the 'trace' routine recommended by the
> parsing and execution page to get definitive answers
> on some of these questions.
So far, I've been avoiding using the trace facility.
I'm interested in J because I prefer working in terms
of abstractions, so I'm inclined to focus on
conceptualizing the relationships and transformations,
rather than attending to transitory machine states.
Perhaps the time has come, however, to add tracing to
my studies.
I've certainly become convinced, from your recent
post, that I must attend to parsing-and-execution. I'm
beginning to see the magnitude of what K. E. Iverson
mentioned in 'A Personal View of APL' when he wrote:
| In programming languages this distinction
| [between word formation and parsing] is
| commonly blurred by combining word formation
| and parsing in a single process characterized
| as "syntax." In J, the word formation and
| parsing are distinct. In its implementation,
| each process is table-driven; the parsing table
| being presented explicitly in the dictionary of
| J, and the rhematic rules being discussed only
| informally.
So, with regard to specifics, on Sat, 30 Jun 2007
14:50:04 -0400 you wrote:
> If you continue evaluating the above phrase,
> you'll find that the verb +/\ gets formed (as
> before), but it will be combined with this fork
> in a hook. But since that's off topic for what
> you are trying to do, I'll not go into details.
In fact, this is on-topic, because what I'm trying to
do is understand how triangularNumber works, and it
turns out that it does form a hook with this as the
left verb, as I thought before I posted anything to
the list.
On Sun Jul 1 13:39:23 HKT 2007 you wrote:
> Note also that verbs of the form u\ use u
> monadically,
That is a breakthrough concept for me. Looking in the
dictionary, I remain wholly unclear as to how to
discover this fact, but your mentioning it greatly
clarifies my efforts to comprehend the verb I wrote
(triangularNumber).
Thank you for the tip (regarding <\) for examination
of what arguments are occurring when the advberb \ is
used.
> ...
> Neither of the above are hooks. But ~ is an
> adverb which always uses its verb dyadically.
What I said was:
TH> the removal of @ or @: in +/[EMAIL PROTECTED] noun
TH> produces a hook, where +/\ is g and someverb
TH> is h
So, going from
+/\@:}.@:i.@:>:
to
+/\}.@:i.@:>:
produces a hook, being:
(+/\) (}.@:i.@:>:) ]
(If I can write it so.)
It was because I know that the left verb in a hook is
dyadic that I examined the dyadic processing of +/\
(only to encounter something that has no relevance to
my use of the same verb, in the same manner!)
It seems that I must put in more parsing-and-execution
study before I fully get how this verb
(tN=:+/\}.@:i.@:>:)
returns a single value, instead of a table, but I
think some key obstacles have been cleared.
Tracy
____________________________________________________________________________________
Choose the right car based on your needs. Check out Yahoo! Autos new Car
Finder tool.
http://autos.yahoo.com/carfinder/
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
