[antlr-interest] Re: Local lookahead depth

[leung13512c]

If the decision change the interpretation of what follows, you may
not be able to delay that decision to the next pass.

Well thought out language probably would have avoid such scenerio
but may be there are some bad guys.

--- In antlr-interest at yahoogroups.com, "John D. Mitchell"
<johnm-antlr at n...> wrote:
> >>>>> "Oliver" == Oliver Zeigermann <oliver at z...> writes:
> >>>>>> John D. Mitchell wrote:
> >>>>>>> "Oliver" == Oliver Zeigermann <oliver at z...> writes:
> 
> >> [...necessity of guarded predicates...]
> 
> [...]
> 
> >>> If you can not set the symbol table, you can not parse a
language, you
> >>> otherwise could. Thus this stuff augments the set of parseable
> >>> languages with ANTLR.
> 
> >> You still haven't shown how the "guarded predicates" are
*necessary* to
> >> understanding that type of language.
> 
> > Ooops, I though I had shown necessity?! Where is the missing piece?
> 
> That's my question. :-) All that you've shown is that using "guarded
> predicates" helps you parse that language *only* using a lexer and a
> parser.  You haven't shown how that increases the recognition power of a
> complete translator has been increased.
> 
> As I've noted, you can use additional stages (e.g., antlr-based tree
> transformers) to more cleanly solve your example language problem.
> 
> Take care,
> 	John


 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/