On Oct 23, 2006, at 2:34 PM, Bryan Ford wrote:
I hope sometime soon Terence will find time to write up the very
interesting
analysis algorithms he's been implementing in ANTLR v3 too! [hint,
hint] ;)
Hi Bryan,
Yeah, I was all set to do so including statistics from random
grammars around the web (ANTLR knows how to report home to the mother
ship), but the standards for a PLDI paper would require me like 6
months of hard work with low probability of acceptance. I'd rather
just describe the algorithm and all the "landmines" to look out for
and some preliminary stats; highest expected value from working on
the book / tool instead. Perhaps a sigplan notices thing, just to
timestamp and get credit for it. My skills don't lend themselves
towards academic papers in the language area I've discovered. Or, I
should say that the academic community (this group exempted) has
shown a lack of interest in my work--only worker bees seem
interested...my motivation fell apart recently for publications.
Sorry. LL(*) really is groovy. It provides a good practical mix of
CFG/PEG and provides a unified view of lexing, parsing, and tree
parsing. I expect that this hybrid or similar will become the
dominant strategy over the next few years.
I will eventually write it up for sure...perhaps we should have a PEG
workshop and then I could just give a presentation to you folks? I
might be able to line that up at Univ. of San Francisco.
Terence
_______________________________________________
PEG mailing list
PEG@lists.csail.mit.edu
https://lists.csail.mit.edu/mailman/listinfo/peg
