I've never had any problems with StandardAnalyzer. Could you perhaps post the code snippet that causes this? I'm wondering about things like what JVM you're using, what your classpath looks like, what Lucene version you're using etc.
Erick On 1/17/07, silegav_k <[EMAIL PROTECTED]> wrote:
I have used the same analyzer before and had no problems at all. The only difference is that I used it to search through full documents and not dictionary-like data. I also use the same analyzer in indexing and in searching, so this must not be the problem. I just tried the StandardAnalyzer as you correctly guided me but for now the only thing I got is the following exeption java.lang.NoSuchMethodError: org.apache.lucene.analysis.StopFilter.makeStopSet ([Ljava/lang/String;)Ljava/util/Set; The PerFieldAnalyzerWrapper and Luke I think are going to be extremelly helpfull. I wil check them later when I will have the time. I will try the snowball anayzer in order to get a proper English Stemmer. Stemming english words is my other problem. I simply cannot stem english words. But this is another problem that I postponed for the next stage. If you come up with an idea of what goes wrong, please post it! Thank you! Vagelis Erick Erickson wrote: > > What analyzer are you using when you *index*? Just as the analyzer you use > when you query breaks up the query string, the analyzer you use when you > index breaks up the indexing stream. You can easily get unexpected results > when you use one analyzer for indexing and another for parsing your query. > > I'd recommend a couple of things. > > 1> just use the StandardAnalyzer first. When you start getting expected > results, substitute in your custom analyzer. That way you can deal with > one > new thing at a time. > > 2> get a copy of Luke (google lucene luke). It lets you examine your index > and see if the things you *think* are in the index actually *are*. It also > lets you submit queries using various analyzers and see what is produced > for > queries. I don't know if you can plug in your own custom one though.... > > Whenever I have this kind of problem, it almost always turns out to be an > issue with analyzers not doing what I *think* they're doing, or using the > wrong analyzer when indexing or searching or..... > > By the way, you can easily use different analyzers on different fields, > See > PerFieldAnalyzerWrapper. > > Finally, the Snowball analyzer also does stemming, and I'd always prefer a > stock analyzer to a custom one if it does what I want. You might want to > take a look at it if you haven't already..... > > Hope this helps! > Erick > -- View this message in context: http://www.nabble.com/only-one-search-result-tf3024628.html#a8406325 Sent from the Lucene - Java Users mailing list archive at Nabble.com. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
