[antlr-interest] replacing NoViableAltException with exception with less stack

[Randall R Schulz]

On Monday 10 September 2007 12:43, Jochen Theodorou wrote:
> Hi,
>
> we (Groovy team) have currently several performance problems with our
> grammar. The resulting Parser is a bit too slow for our needs and we
> are now looking for a way to make it faster. The grammar is a big and
> complicated beast and I am sure there are many places we can make it
> better.
>
> But one thing that would speed up the whole thing a little is to
> remove that stack trace used by the NoViableAltException.

Do I understand you to be saying that the grammar used by Groovy 
routinely produces these exceptions even for grammatically acceptable 
input? If so, why? If not, then it cannot be a source of excessive 
overhead.


> ...
>
> Note: atm antlr 3.0 is no option

Eh? Why not?


> thanks, blackdrag


Randall Schulz