[antlr-interest] clarification on tree proposal

Terence Parr parrt at cs.usfca.edu
Tue Feb 8 09:59:36 PST 2005


Howdy,

Just in case I've not been clear on the tree proposal stuff, that blog 
entry is what I am proposing to implement, but the ideas are *not* 
exclusively mine.  I've been working with Loring Craymer, Monty 
Zukowski, Ric Klaren, and John Mitchell very closely over the years and 
I have been greatly influenced by their ideas.  Further, ANTLR users 
(and particularly the ANTLR2004 workshop attendees) continuously send 
me good ideas.  I use whatever I think is useful and can fit in the 
"concept" I'm shooting for.  I know that Matthew Ford sent me a big 
list of ideas for the tree stuff, which I've read and, hence, been 
influenced by.   Another case in point: John Mitchell just proposed 
label+=element to better indicate you intend to work with all values 
instead of label=element, which means "last element matched."  Love 
that.  I will add to the proposal when I get a few spare cycles.

Regarding Loring's 2.8experimental, I hope to use as many ideas as 
possible in my implementation for v3.0.  Certainly, he has some great 
ideas, which I've already included in my proposal.  As I learn more, I 
may grab even more! :)  We should get some good experimental data on 
the ideas as people start to play with it.  I will watch closely.

Anyway, just wanted to be clear about things.  Keep those ideas coming 
folks!  ANTLR has become popular because I try to provide what you ask 
for! ;)

Thanks,
Ter
PS	the same collaborative effort went into the attribute mechanism.
--
CS Professor & Grad Director, University of San Francisco
Creator, ANTLR Parser Generator, http://www.antlr.org
Cofounder, http://www.jguru.com





More information about the antlr-interest mailing list