a. As suggested in E. Parsing and Execution, try:
   load 'trace'
   trace 'f=: >./ , <./'

b. The stack is initialized with 4 markers, so that
the parser never has to deal with a stack with less
than 4 elements.



----- Original Message -----
From: June Kim <[EMAIL PROTECTED]>
Date: Friday, January 19, 2007 5:04 am
Subject: Re: [Jprogramming] Parsing rules

> 2007/1/19, Miller, Raul D <[EMAIL PROTECTED]>:
> > June Kim wrote:
> > > At the near bottom of E. Parsing and Execution at
> > > system\extras\help\dictionary\dicte.htm , there is the
> > > precise rule for parsing and execution for J sentences.
> > >
> > > However, I don't understand its effects when applied to
> > > the following sentence.
> > >
> > > f=: >./ , <./
> >
> > Based on the II.E rules:
> >
> > ADJ  NONE NONE NONE      /
> > VERB ADJ  NONE NONE      <.   /
> > VERB VERB ADJ  NONE 3    ,    <.   /
> > VERB VERB NONE NONE      ,    <./
> > ADJ  VERB VERB NONE      /    ,    <./
> > VERB ADJ  VERB VERB      >.   /    ,    <./
> > ASGN VERB ADJ  VERB 3    =:   >.   /    ,    <./
> > ASGN VERB VERB VERB 5    =:   >./  ,    <./
> > ASGN VERB NONE NONE      =:   >./,<./
> > NAME ASGN VERB NONE 7    f    =:   >./,<./
> > VERB NONE NONE NONE .    >./,<./
> >
> > I can imagine several areas of confusion, here:
> >
> > The linear representations of the derived verbs look
> > like the unparsed characters.
> >
> > The stack sometimes contains less than four elements.
> >
> > The stack sometimes contains more than four elements.
> >
> > That said, if some step is not clear to you, please
> > ask a specific question about it.
> 
> Very clear. Thanks.
> 
> It seems like my assumptions fit with the model your describe. My
> initial difficulty was at realizing ANY could match "NONE"(I couln't
> think NONE is something(nothing) existing at all).
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to