[antlr-interest] Re: updated "antlr 2 bashing list"
Terence Parr
parrt at cs.usfca.edu
Tue Mar 9 17:22:25 PST 2004
On Mar 9, 2004, at 4:51 PM, lgcraymer wrote:
> Ter--
>
> I hate to say it, but I think that that's a cop out. Building
> exception objects and stack frames costs, and there is no point when
> in guessing mode.
This never happens unless you fail and is a minor cost compared to the
cost of nested predicates, right?
> They are also obnoxious when debugging (especially
> C++)
Doesn't bother me in Java, but other languages might be a drag.
I can investigate not using exceptions to see what the code looks like.
Actually somebody who has used JavaCC could tell us what it looks like
;) They do not use exceptions.
> and hamper code generation for languages which do not have
> exceptions. It would be better to have a "synpredFailed flag as part
> of the generated classes-the flag will always be in cache, so is not a
> performance issue--or to have a return flag from rules and wrap rule
> invocations (foo) in
> if (foo() == false) return;
>
> and add
> if (guessing)
> return false;
> else {
> <throw exception>
> }
>
> around the <throw exception>s in ANTLR. Come to think of it,
> implementing this with the current ANTLR runtime would be only about
> 10 lines of code or so (maybe a little more--I think that each
> exception would have to be fixed separately).
Uh...how you gonna redo return values in 10 lines ;)
Ter
--
Professor Comp. Sci., University of San Francisco
Creator, ANTLR Parser Generator, http://www.antlr.org
Cofounder, http://www.jguru.com
Cofounder, http://www.knowspam.net enjoy email again!
Cofounder, http://www.peerscope.com pure link sharing
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/antlr-interest/
<*> To unsubscribe from this group, send an email to:
antlr-interest-unsubscribe at yahoogroups.com
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
More information about the antlr-interest
mailing list