[antlr-interest] RFC: slight change in ! semantics

Ruslan Zasukhin sunshine at public.kherson.ua
Sat Oct 6 01:41:17 PDT 2001

on 10/6/01 1:02, Matthew Ford at Matthew.Ford at forward.com.au wrote:

> I still prefer version 2)
> But I as I understand it  3) is just 2) with a lot of extra code for all the
> non-labelled  IDENT's etc.

> I would have thought that Ruslan Zasukhin would have been jumping up and
> down about this extra code and be voting for 2).

Good note, Mathew...

but...this extra code will be in ANTLR code generator,
    NOT in generated code of parser.

This make difference.

We run ANTLR parser only on our side, I think nobody care if generation of
code will take 2 or 3 seconds.

But our parser will be run by thousands of our users million of times.
So important is to get fast generated PARSER.

And again, on my point of view, ideal way now -- 4.
** One develop can set ON optimization and then he need label all tokens as
you want (way 2)
** Other developer can use OFF, and still work without changes (lazy way 1).

NOTE: way 4 in fact is combination of ways 1 and 2.

Best regards,
Ruslan Zasukhin


e-mail: ruslan at paradigmasoft.com
web   : http://www.paradigmasoft.com

To subscribe to the Valentina mail list
send a letter to valentina-on at lists.macserve.net


Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/ 

More information about the antlr-interest mailing list