Re: Query scoring

2009-04-18 Thread liat oren
Hi, You can see the output below: 1.0 * (this is the score) 4817674 5664274 6361333 (query) 4817674 5664274 6361333 (doc 0) 0.0 = (NON-MATCH) product of: 0.0 = (NON-MATCH) sum of: 0.0 = coord(0/3) 0.13811989 * 4817674

Re: Indexing Complex XML

2009-04-18 Thread Paul Libbrecht
daniel, have a look at solr DIH, it has prebuilt tools to do just that. http://wiki.apache.org/solr/DataImportHandler This bases on solr which is a web-application that bases on lucene. It does not need imperatively to be run as a web application though, it can be embedded. paul Le 18-avr

RE: Indexing Complex XML

2009-04-18 Thread Digy
doc.add(new Field("authors", "name1 surname1 name2 surmane2", StoreOption, IndexOption); So you can make a search like authors:"name1 surname1" (Disadvantage: you will also get result with a search like authors:"surname1 name2" ) DIGY -Original Message- From: Daniel Susanto [ma

Re: Indexing Complex XML

2009-04-18 Thread Daniel Susanto
Thanks Erick, In more complex xml I mean, for example this xml: Lucene Book Book author 1 Book author 2 Book for Lucene Lucene Book 2 Book 2 author 1 Book 2 author 2 Book 2 for Lucene for each 'book' node is handled by one Document rite? and now how to handle the 'authors' nod

Re: NativeFSLockFactory still leaving lock files around on abnormal termination

2009-04-18 Thread Michael McCandless
This is normal: the file will be left in the directory, but, the lock will not be held (the OS releases it, but does not delete the file). I think the javadocs state this somewhere. Mike On Sat, Apr 18, 2009 at 11:44 AM, Newman, Billy wrote: > So I have recently switched to using NativeFSLockFac

Re: Indexing Complex XML

2009-04-18 Thread Erick Erickson
Lucene is an *engine*, not an application. *You* have to process the XML, decide what the structure of your index is and index the data. There are many XML parser options, this is just straight Java code. You'll decide what's relevant, and add the contents of the relevant elements to a Lucene docum

Re: Taxonomy in Lucene

2009-04-18 Thread Karsten F.
Hi John, I intended to compare xtf with hierarchical facet browsing in browseengine (selection expansion). I found PathFacetCountCollector/PathFacetHandler#getFacetsForPath, and I think that the implementation in xtf has a lot of advantages. So I suggest you to reuse the xtf-source for that (Gro

NativeFSLockFactory still leaving lock files around on abnormal termination

2009-04-18 Thread Newman, Billy
So I have recently switched to using NativeFSLockFactory so that when my jboss instance is shutdown or comes down abnormally it does not leave lock files around. My problem is that it is not working. I have put in debug to be sure that I am really using NativeFSLockFactory. I have tested this

Indexing Complex XML

2009-04-18 Thread daniel susanto
Hi, I need advise or example to index complex XML file, I mean the XML note just in one level node but more than one. for example indexing rss or atom. thx b4. Daniel Susanto http://susantodaniel.wordpress.com

Re: Faceting, Sort and DocIDSet

2009-04-18 Thread Karsten F.
Hi Dave, searching and sorting in lucene are two separate functions (if you not want to sort by relevance). You will not loss performance if you first search with BitSet as HitCollector and then sort the result by DateField. But more easy is to extend TopFieldDocCollector/TopFieldCollector to a C