|16 bits lexers... email@example.com (Darius Blasband) (1997-05-04)|
|Re: 16 bits lexers... firstname.lastname@example.org (Scott Stanchfield) (1997-05-08)|
|Re: 16 bits lexers... email@example.com (Jerry Leichter) (1997-05-08)|
|From:||Scott Stanchfield <firstname.lastname@example.org>|
|Date:||8 May 1997 21:26:50 -0400|
Darius Blasband wrote:
> There have been quite some arguments about how to implement scanners
> dealing with 16 bits (unicode) transitions. One of the major issues
Another approach is to do what Terence Parr has done for PCCTS 2.0 --
the scanner is recursive-descent, just like the parser. No tables.
It will eventually have some nice auto-inlining optimizations, and the
generated code is actually pretty similar to what you might write by
hand for a scanner. It's written in Java and handles Unicode...
Take a peek at
Santa Cruz, CA
Return to the
Search the comp.compilers archives again.