lucene-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Chris Currens <chris.curr...@discover-e-legal.com>
Subject Expected Behavior from QueryParser and Standard Analyzer with Version.LUCENE_*
Date Mon, 09 May 2011 16:24:50 GMT
Hello,

I have some questions about what kind of behavior is expected when passing
Version.LUCENE_24/29/30 to QueryParser and the StandardAnalyzer when parsing
a query.  I know that passing the Version to the constructors make Lucene
act that like version, with all features and bugs intact.  The behavior I
expect (and am getting from these versions, in both Java and .NET Lucene) is
when passing a query to parse that is in quotes and includes stopwords,
would be a lowercase string in quotes, with the stopwords removed.  ex.
"Calendar item as Msg" -> "calendar item msg".

I'm getting different results from pylucene, which just invokes a lucene
jar.  It parses "Calendar item as Msg" -> "calendar item ? msg", which
causes the query to fail.   People from the pylucene project want to know
the expected behavior, as they recall it behaving differently in past
versions.  I see no reason that the behavior would change from using the
same version of lucene in pylucene vs straight java.

Your answer is greatly appreciated.

Thanks,
Christopher

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message