Hi all !
This thread fills up with a very strange discussion, but don't miss Jose
Mario Quintanas interesting post.
http://www.jsoftware.com/pipermail/programming/2017-October/049146.html
Opinions are welcome!
Cheers,
Erling Hellenäs
Den 2017-10-13 kl. 13:06, skrev Erling Hellenäs:
No, as I understand it it looks ahead one. I am sure you are right on
that. And when the lookahead item to the left is not a conjunction
what happens then? The NOUN VERB NOUN combination is not executed or
composed? Or you call the NOUN VERB NOUN combination a bident as Raul
states? /Erling
Den 2017-10-13 kl. 12:55, skrev bill lam:
Note the AVN instead of CAVN.
If it is a C then the next N would be its argument and executed first
because a conjunction has a higher binding power than a verb, ie J
execution is not necessarily strictly from right to left even if the
total
absence of parentheses. This is my understanding but of course I may be
wrong.
On Oct 13, 2017 6:36 PM, "Erling Hellenäs" <[email protected]>
wrote:
Hi all!
http://www.jsoftware.com/help/dictionary/dicte.htm
"3. A verb is applied dyadically if possible; that is, if preceded
by a
noun that is not itself the right argument of a conjunction."
"EDGE+AVN NOUN VERB NOUN 2 Dyad"
/Erling
Den 2017-10-13 kl. 11:23, skrev bill lam:
The J parser is unambiguously defined by the online webpage dicte as
pointed by Raul, or page 80 in my previous image. It is quite simple.
Please state which part you couldn't understand.
I am curious, you seemed to have written a simple version of J
interpreter,
how was it possible without knowing the J execution stack model in
page 80?
On Oct 13, 2017 4:53 PM, "Erling Hellenäs" <[email protected]>
wrote:
Hi all!
As I see it there are two ways to handle a dyadic verb in a parser.
Either
you react on the trident N0 V1 N2 or you partially execute V0 N1. In
this
case the product must be a new monadic verb. This new monadic verb
is then
handled further in the parse process. However, I can not see any verb
results of V0 N1. This means I don't understand how the parser could
execute a dyadic verb(like in explicit J) or how it could compose the
dyadic verb with previous compositions (like in tacit J).
According to my analysis handling N0 V1 N2 as a trident would be
much more
efficient, since the other alternative would be to execute each verb
twice
or compose it for later execution twice.
It seems in J partial execution of dyadic verbs is not possible
since the
monadic form of the verb would then be executed and the left noun
argument
would cause a syntax error.
Maybe Bill or Raul can in clear terms describe how the parser handles
dyadic verb application so that we can put an end to this very
strange side
discussion?
Cheers,
Erling Hellenäs
Den 2017-10-13 kl. 04:59, skrev Raul Miller:
Or it looks like I was incorrrect, in terms of the thinking that went
behind it.
Even though the noun train would not handled as a trident by the
parser, it looks like Ken had that in mind as a part of the structure
of the implementation.
Thanks,
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
