I've been to hell and back to make it faster, but that's it, shift-reduce can't be much faster. Unfortunately, finding errors in text makes it three times slower. At least it has linear parsing time.
i've just fixed some ugly bug about unspecified tokens. sorry for inconvenience.
oh, did i mention that the code is free for any use? just don't sue me, that's all
*Nothing against JS. I just can't see it being easy to write a parser in.
A new version of Moony Parser is out. This time it implements "Earley" parser. Also handles all CFGs. It is up to 100 times faster than previous version in some cases. There should be less bugs because the code is simpler.