[antlr-interest] v3.0b5: input size problem

Randall R Schulz rschulz at sonic.net
Fri Nov 24 18:43:12 PST 2006


Paul,

On Friday 24 November 2006 17:07, Paul Shaffer wrote:
> Sorry, I see this stopping is actually due to my parsing errors. As I
> fix the errors I am able to parse farther into the input text, I am
> up to line 1600 or so now. But there are no errors being reported
> when it stops. I wish I knew how to get errors to show up-- any
> suggestions appreciated.

Are you sure you're invoking the JVM that runs the ANTLR-generated 
parser with enough RAM? It's typical for execution to appear to grind 
slowly to a halt as more and more of the CPU time goes to running the 
garbage collector.

As Terence pointed out, by default ANTLR parsers process the entire 
input, transforming it into a linear sequence of tokens that, in 
addition to the input text itself, contains a list of Token instances 
that partition that input.

If you don't have enough RAM allocated to the JVM, then it will fail at 
some point during analysis of the input.


Randall Schulz


More information about the antlr-interest mailing list