Or, slightly more precisely, the matching works like this VERB: ! matches EDGE+AVN NOUN: 4 matches VERB+NOUN ADV: / matchs ADV ANY: empty is a match
This is the adverb rule which consumes the two middle items, and of course 4/ is a domain error in J. Thanks, -- Raul On Fri, Oct 1, 2021 at 11:34 AM Henry Rich <[email protected]> wrote: > > ! 4 / matches line 3, EDGE+AVN N+V ANY > > It executes / with operand 4 which is a domain error > > thus the spaces before the / in the error line. > > Henry Rich > > On 10/1/2021 11:03 AM, David Lambert wrote: > > I'm having some difficulty understanding the parse table. What means > > "ANY"? The first rule, if ANY means "arbitrary token" then I expect > > the error as t_z=. (24) (/) > > > > <lang LaTeX> > > > > \newcommand{\be}{1}{\bold{\textit{#1}}} > > > > EDGE \be{VERB} \be{NOUN} ANY 0 Monad > > > > </lang> > > > > <lang j> > > > > trace '! 4 /' > > --------------- 3 Adverb ----- > > 4 > > / > > |domain error: executet > > | t_z=.(4) (/) > > > > </lang> > > > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > -- > This email has been checked for viruses by AVG. > https://www.avg.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
