>
> I think this strategy would be completely infeasible for a real grammar.  
>

Perhaps you meant that for the grammar you have to contend with, but I 
still disagree. There are 2 courses of action here too (which is always 
better than having 1 or none):

1: Extend the original grammar so Marpa parses everything with 1 grammar

2: Have a second recognizer on stand-by, with it's own grammar for the 
strings which trigger rejection event in your original code.

I've done both, which I why I seriously recommend them to you.

For an example of the former, see GraphViz2::Marpa, and search for the 
HTML-specific grammar. It's the second - external - grammar of the 2. 
 Search GraphViz2/Marpa.pm for the string 'bnf4html'.

-- 
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