On Friday 20 May 2005 17:20, Terry Steichen wrote: > Paul, > > Could you flesh out the implementation you describe below with some code > or pseudocode?
You can start from this: http://issues.apache.org/bugzilla/show_bug.cgi?id=34331 and use code from the method in src/java org.surround.search.SpanNearClauseFactory.makeSpanNearClause() and from the classes SrndPrefixQuery and SimpleTerm. But it's probably easier to start from the javadocs of the org.apache.lucene.search.span package and the java code from the 1.4.3 PrefixQuery. Have a nice weekend, Paul Elschot > Regards, > > Terry > > > Paul Elschot wrote: > > >On Friday 20 May 2005 11:30, Stanislav Jordanov wrote: > > > > > >>Is there a Lucene Query (or something that will do a job) like: > >>"Star Wars tri*" > >> > >>that will match all docs containing a 3 word phrase: 'Star' followed by > >>'Wars' followed by a word starting with 'tri'. > >> > >>I.e. the above query will match both "Star Wars trilogy" and "Star Wars > >>triumph". > >> > >> > > > >You'll need an ordered SpanNearQuery over the following: > >- SpanTermQuery for "Star" > >- SpanTermQuery for "Wars" > >- SpanOrQuery over all SpanTermQuery's for terms matching tri*. > > > >The last one should be a SpanPrefixQuery, but that one is not > >available. Have a look in PrefixQuery.rewrite() on how to find all > >terms matching tri*, it's fairly straightforward. > > > >Regards, > >Paul Elschot > > > > > >--------------------------------------------------------------------- > >To unsubscribe, e-mail: [EMAIL PROTECTED] > >For additional commands, e-mail: [EMAIL PROTECTED] > > > > > > > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
