Re: Lexer capable of 16-bit characters

Nope -- 2.4 doesn't support UNICODE yet.


We've made the following changes (and a few others) in 2.4.0:


* Added the first release of ParseView, my visual debugger
    (check it out at http://java.magelang.com/~thetick/parseview)


* Set it up to support Readers instead of just InputStreams (a good step
toward UNICODE)


* Fixed the output and input (of antlr.Tool, the parser generator) to
properly handle newlines on all platforms


* added a -o option to specify an output dir




The problem left with UNICODE is that grammar analysis will blow up.
Everything is based on small bitsets to match characters. Works great for
ASCII ;)


We've come up with some very cool ways to implement "imaginary bitsets" to
deal with things like ~'a', but it's a matter of Ter getting a bit of free
time to implement them.


ANTLR (including ParseView) is totally free for whatever use anyone wants
(other than drumming up a silly lawsuit against Ter). We only ask that if
you use it you put a little ref about it in your docs somewhere.


You can get ANTLR & info at http://www.antlr.org (ANTLR 2.3.0)


NOTE: 2.4.0 is available as a "preview" (announced in
comp.compilers.tools.pccts) at
http://www.antlr.org/nirvana/antlr-2.4.0b2.zip THIS IS STILL A BETA!


--
-- Scott


Scott Stanchfield Santa Cruz, CA USA
    thetick@magelang.com http://java.magelang.com/~thetick


Gary D. Duzan wrote
> I believe the upcoming 2.4 version of ANTLR is supposed to include
>full Unicode support. Check out < http://www.antlr.org/ >.
--

Return to the comp.compilers page.
Search the comp.compilers archives again.