Can't help thinking that the recognizer's method, which reads input and
returns a parse if there is one or throws an exception otherwise might (or
even must) well be called parse().



On Mon, Aug 18, 2014 at 8:31 PM, Jeffrey Kegler <
[email protected]> wrote:

>  I've just shadowpasted a first cut at the new doit() method
> <http://scsys.co.uk:8002/416488>, as discussed in this group, so that
> those interested can look it over.  You'll have to hunt for it in the file
> a bit -- the actual method is buried deep in a test program.
>
> I settled on doit() as the name.  process() was the best alternative, but
> 'process' is just a Latin-ism for 'doit'.  In this first cut, doit() is in
> the 'main' namespace -- when finished it will be a Marpa::R2::Scanless::R
> method.
>
> doit() reads an input and on success returns a ref to the only parse.  On
> any kind of failure, it throws an exception.  Very importantly, ambiguity
> (more than one parse) is considered an error, and reported as such.  Also,
> if any events cause the read to terminate prematurely, an error is thrown.
>
> The intent is that this is a very simple method, good for beginners and
> for a first cut, and uncomplicated by options.  If you want events, or want
> to treat an ambiguous parse as anything but an error, or want fine control
> of the errors, you should use the read() and value() calls directly.
>
> Thanks, jeffrey
>
> --
> You received this message because you are subscribed to the Google Groups
> "marpa parser" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups 
"marpa parser" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to