[antlr-interest] we can do push/pop error stuff for 4%

[Olivier Dragon]

On Sat, May 27, 2006 at 12:51:29PM -0700, Terence Parr wrote:
> 
> On May 27, 2006, at 12:44 PM, Sohail Somani wrote:
> >Aren't java containers thread safe or something? Might be
> >synchronize(this) or something like that...
> 
> Good point, though I thought that synchronization had gotten much  
> faster these days...

I think it might also have to do with fact that java.util.Stack is
based on Vector, which is a resizeable array with default size 10. If
your grammar uses more than 10 elements it's going to do a lot of
resizing.

If you're interested in seeing the possible effect of either
synchronisation or resizing, build a stack class using
java.util.ArrayList which is resizeable but not synchronised. I presume
you build your stack with a fixed size array?

Do you know ahead of time the maximum size of the Stack based on the
grammar?

-Olivier

-- 
          __-/|    ? ?     |\-__
     __--/  /  \   (^^)   /  \  \--__
  _-/   /   /  /\ / ( )  /\  \   \   \-_
 /  /   /  /  /  (   ^^ ~  \  \  \   \  \
 / Oli Dragon    ( dragonoe at mcmaster.ca \
/  B.Eng. Sfwr   ( dragon.homelinux.org  \
/  /  /    /__--_ (   ) __--__\    \  \  \
|  /  /  _/        \_ \_       \_  \  \  |
 \/  / _/            \_ \_       \_ \  \/
  \_/ /                -\_\        \ \_/
    \/                    )         \/
                        *~
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
Url : http://www.antlr.org/pipermail/antlr-interest/attachments/20060527/f777dfe6/attachment.bin