in general most stuff will be ok (code just doesnt mess with the surrogate pairs)
simple example: things like lowercase filter are not, and lowercasetokenizer, well, can't fix that since its super chartokenizer, works on char. i hope this doesnt sound bad, but in general the transition is simple for most apps, because they work with String. Lucene works with char a lot, so its a bigger challenge, but I appreciate the good performance :) On Fri, Jun 12, 2009 at 12:45 PM, Michael McCandless<luc...@mikemccandless.com> wrote: > On Fri, Jun 12, 2009 at 9:09 AM, Robert Muir<rcm...@gmail.com> wrote: >>> >>> 3.0 will then be a mechanical release: "simply" remove all deprecated >>> APIs, fix all core, contrib, tests to not use those APIs, release, and >>> immediately begin accepting 1.5 patches. >>> >> >> personally, I wonder if anyone but me is interested in correcting the >> issue that unicode character no longer fits in java 'char' in java 1.5 >> :) >> >> maybe I am the only one, but i still think its correctness issue..., >> and will require changing some apis, even if its just char->int > > I think this is important. > > But it's a big change right? EG would we need to change tokens to be > int not char based? > > Also, can't this happen after 3.0? > > Mike > > --------------------------------------------------------------------- > To unsubscribe, e-mail: java-dev-unsubscr...@lucene.apache.org > For additional commands, e-mail: java-dev-h...@lucene.apache.org > > -- Robert Muir rcm...@gmail.com --------------------------------------------------------------------- To unsubscribe, e-mail: java-dev-unsubscr...@lucene.apache.org For additional commands, e-mail: java-dev-h...@lucene.apache.org
