Just read that now.  Very useful stuff.  Thank you.

A new mystery though...

EDGE+AVN VERB+NOUN VERB VERB  5 Trident (Fork)
EDGE CAVN CAVN ANY  6 Bident (Hook/Adverb)

(N V V) can be a fork? -- through testing I see that
this makes what I thought should be called a dyadic
hook.  I presume the name doesn't matter so much, just
that the pattern is parsed as something.

For line 6,
EDGE N N ANY -> syntax error
EDGE C A ANY -> syntax error
EDGE C C ANY -> syntax error
and probably more.... (NV) is syntax error
Does this pattern only really match?
EDGE A A ANY -> adverb
EDGE V V ANY -> hook



--- Henry Rich <[EMAIL PROTECTED]> wrote:

> I did a decnt job of explaining this in J for C
> Programmers, in
> the chapters towards the end on Parsing.
> 
> It turns out to be pretty hard to say what a verb
> phrase is.
> The only way I could come up with makes reference to
> the
> parsing rules.
> 
> Henry Rich 
> 
> > -----Original Message-----
> > From: [EMAIL PROTECTED] 
> > [mailto:[EMAIL PROTECTED] On
> Behalf Of p j
> > Sent: Friday, March 24, 2006 9:43 PM
> > To: Programming forum
> > Subject: [Jprogramming] Parsing conjunctions
> > 
> >    (-: *:)@+: 2
> > 0
> >    -: *:@+: 2
> > 8
> >    -: (*:)@+: 2
> > 8
> >    +:@-: (*:)@+: 2
> > 16
> >    +:@-: *:@+: 2
> > 16
> >    +:@(-: *:@+:) 2
> > 0
> >    +:@-: (*:@+: 2)
> > 16
> > 
> > >From the dictionary,  conjunctions have as left
> > argument "the entire verb phrase that precedes
> it."
> >    p=: 2 : 'u'
> >    +:@-: *:@+: p 1
> > +:@-: *:@+:
> > That doesn't mean very much, because the
> relationship
> > of the entire verb phrase is dependent upon the
> single
> > word (or paren'd entity) that is to the left of
> the
> > conjunction.
> > 
> >    +:@-: 2 + *:@+: 2
> > 18
> >    +:@-: 2 + (*:@+:) 2
> > 18
> > 
> > >From the above examples, it looks a lot like you
> could
> > think of conjunctions as binding with the word on
> its
> > left (which will then be processed by the rest of
> the
> > sentence)
> > 
> > Amazingly,
> >    +:@-: *:@(+: p 1)
> > +:@-: *:@+:
> >    +:@-: 2 + p + +  
> > +:@-: (2 + +)
> >    +:@-: 2 (+ p +) +  
> > +:@-: (2 + +)
> > 
> > Can you say that for any conjunction c, sentences
> s
> > and words w:
> > s3 (s2)c(s1) -: s3 ((s2)c(s1))
> > s3 w2 c w1 -: s3 (w2 c w1)
> > 
> > I'm very surprised that there seems no way to
> limit
> > the left argument to a conjunction by
> parenthesising.
> > 
> > Here's another example that better illustrates the
> > concept that conjunctions really left-bind with
> one
> > word:
> >    -/  +/"(0) 3 2 3
> > 4
> >    -/"0  +/"(0) 3 2 3
> > 3 2 3
> >    (shouldn't rank be setting all verbs on the
> left
> > side?)
> > 
> > what semantics make that meaningful?
> > 
> > __________________________________________________
> > Do You Yahoo!?
> > Tired of spam?  Yahoo! Mail has the best spam
> protection around 
> > http://mail.yahoo.com 
> >
>
----------------------------------------------------------------------
> > For information about J forums see 
> > http://www.jsoftware.com/forums.htm
> 
>
----------------------------------------------------------------------
> For information about J forums see
> http://www.jsoftware.com/forums.htm
> 


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to