[antlr-interest] Unicode Scanners and java.g

Jeremy Rayner jeremy.rayner at gmail.com
Thu Mar 31 11:05:52 PST 2005


Hiya,
  In lieu of trawling through the historical copies of java.g, what are the
reasons in Michael Studman's latest copy for having the scanner use
the following...

charVocabulary='\u0003'..'\u7FFE'

I understand of course that the top end of this (8bit unicode I assume) is not
treading on the toes of EOF (i.e. \u7FFF as per
http://www.antlr.org/doc/lexer.html#unicode )

But why are the 3 characters '\u0000', '\u0001' and '\u0002' left out?
What are the design decisions behind that?


Thanks for a great tool Ter, Groovy is on the cusp of releasing it's
first antlr based groovy compiler very soon now, yay.

Jez.
-- 
http://groovy.codehaus.org
http://javanicus.com/blog2


More information about the antlr-interest mailing list