On Mon, Apr 11, 2011 at 04:54:40PM -0600, Alan Post wrote:
> On Tue, Apr 12, 2011 at 08:34:49AM +1000, Erik de Castro Lopo wrote:
> > Yury Euceda wrote:
> >
> > > Well, I have this option in my tool I developed. But I defined my own way
> > > due my
> > > own need (And well, I didn't know anything about this but I supposed to
> > > be
> > > necesary
> > > and I invented my own way not knowing the other notation)
> > > for example:
> > >
> > > a>>3
> > >
> > > defines that a MUST appear three times ---> aaa is accepted
> > >
> > > and
> > >
> > > a<<3
> > >
> > > defines that a can appear 0, 1, 2 or 3 times ---> epsilon, a , aa , aaa
> > > are
> > > accepted
> > >
> > > I would like to hear about your comments for my notation.
> >
> > I think that notation is far less clear than the regular expression
> > notation that has been used in lexers for 30+ years. Ie
> >
> > a{0,3}
> >
> > Erik
>
> I know you didn't reply to me. I'm the original author of this
> thread, and since it is alive again, I did go with the {n,m}
> notation exactly as presented by another user on this list.
>
> I made this one of my atomic operators so I could push rule
> counting down into what amounts to my inner loop. I was able to
> simplify a couple of my existing grammars using this form as well.
>
> -Alan
I fear I might have hastily worded the preface to my reply, please
allow me to try again:
Erik, I'm not Yury, who you replied to; I was involved earlier in
this thread by opening it with a question about {m,n} rules in PEG,
and I realized I had not followed up with the syntax I used to
implement this rule. Seeing this thread I wanted to follow up to
acknowledge I used an implementation proposed by another member of
this list, Michaeljohn Clement.
I hope I didn't cause confusion by replying to this thread rather
than to Michaeljohn Clement's thread.
-Alan
--
.i ma'a lo bradi ku penmi gi'e du
_______________________________________________
PEG mailing list
PEG@lists.csail.mit.edu
https://lists.csail.mit.edu/mailman/listinfo/peg
