lucene-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Pasha Bizhan" <fc...@ok.ru>
Subject RE: Using Russian analyzer in Luke
Date Sun, 25 Jan 2004 22:36:11 GMT
Hi, 

> From: Erik Hatcher [mailto:erik@ehatchersolutions.com] 
> 
> On Jan 25, 2004, at 2:53 PM, Pasha Bizhan wrote:
> > Hi,
> >
> > I'm not sure that's rightly. Because Russian unicode charset, KOI 
> > charset and win1251 charset is equal in use. May be unicode 
> charset is 
> > less common.
> > I guess so Russian Analyser hasn't no-arg constructor.
> 
> Pasha - my apologies, but I'm not following what you mean 
> here.  Is there an issue with the change I made to the 
> RussianAnalyzer?

Sorry for my English.

All russian charset is equal. 
I can't choice default charset for no-arg constructor.

My code is only example  and RussianCharsets.RussianUnicode too.
We use RussianCharsets.CP1251. But other people can use other charset.

> public RussianAnalyzer() {
> 	this.charset = RussianCharsets.UnicodeRussian;
> 	stoptable =
StopFilter.makeStopTable(makeStopWords(RussianCharsets.UnicodeRussian));
> }

I think that Russian Analyser must not has no-arg constructor.
The choise of default charset is not evident.
The user must specify charset.

I thin that original problem is Luke's problem, not Russian Analyser.

Do you understand me?

 > My preference, actually, is for Lucene 2.0 to have all of 
> these language-specific analyzers separate from the core.

+1.

Pasha Bizhan
http://www.LuceneDotNet.com


---------------------------------------------------------------------
To unsubscribe, e-mail: lucene-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: lucene-user-help@jakarta.apache.org


Mime
View raw message