Re: The values which compute scores. - Part II

Hi Walt, AFAIK there is no flag guiding scorers to "ignore norms". I guess you could hack a "all norms are 1" behavior by writing something similar to OneNormsReader in org.apache.lucene.demo.SearchFiles. Doron Walt Stoneburner wrote on 01/06/2007 13:45:26: > I've managed to build my own Simila

Re: Two differently sorted result sets from the same search

Carlos Pita wrote on 01/06/2007 13:59:25: > 2) Instead of a TopFieldDocCollector use some > kind of bounded priority queue optimized for > top-N results. With a HitCollector, a Filter > and two of these queues it's easy and efficient > to find both result sets on one simple pass. > Do you know of

Re: Two differently sorted result sets from the same search

Mh, now I found out that latests sdks provide a priority queue too, although unbounded. But, supposing the queue has already top-n elements, I could compare the current heap with the head (O(1)), if it is smaller or equal just discard it, and another way remove the head ((O(log(top-n))) and insert

Two differently sorted result sets from the same search

Hi all, I need to find a couple of result sets at the same time from the same search-criteria. The two sets are sorted according different sort-criteria. From both I need just the few top N results, but anyway, because of business rules, I need to process the entire hit set for the search. Till

The values which compute scores. - Part II

I've managed to build my own Similarity class, plug it in, and use Explain to convince myself that I am, indeed, getting the correct weightings that I desire. My test case documents are yielding precisely the intermediate values needed for alternate scoring. There's just one thing... When I do

Re: question about lucene

If your data is in database, DBSight can get you started easily in 3 minutes! http://wiki.dbsight.com/index.php?title=Create_Lucene_Database_Search_in_3_minutes Highlighter is the part you want when you display the search results. -- Chris Lu - Instant Scalable Full-Text

Re: question about lucene

I suspect the Highlighter (in "contrib" section) is what you're after. There is a patch in JIRA from Mark Miller which adds support for "proper" highlighting of phrases which I promise to take a look at when I get some time (hang in there Mark!) - Original Message From: Will Johnson <[

RE: question about lucene

Solr, which is built on top of lucene and adds highlighting among other features, gets close to what you want. Check out: http://wiki.apache.org/solr/HighlightingParameters - will -Original Message- From: Erick Erickson [mailto:[EMAIL PROTECTED] Sent: Friday, June 01, 2007 8:57 AM To:

Re: question about lucene

Nope. But here's what I think you can do (although I haven't tried this exactly, so caveat emptor). Document doc = new Document(); doc.add("text", line1); doc.add("text", line2); doc.add("text", line3); writer.add(doc); Now, when searching, you can get the document back and String[] lines = doc

RE: question about lucene

Wow, it was fast! Thanks. Do you know about any existing application that is built on top of lucene that provides this functionality? Tanya -Original Message- From: Erick Erickson [mailto:[EMAIL PROTECTED] Sent: Friday, June 01, 2007 7:18 AM To: java-user@lucene.apache.org Subject: Re: q

Re: question about lucene

No. Lucene is an *engine*, not an app that has a lot of stuff built on top of it out of the box. You have to index enough information to figure this out somehow. Best Erick On 6/1/07, Tanya Levshina <[EMAIL PROTECTED]> wrote: Hi, I've just downloaded Lucene, tried demo and looked at the do

Re: Lucene index content

Let's say you indexed the filename in "fname", as in doc.add(new Field("fname", "blah.txt", Field.Store.Yes, Field.Index.TOKENIZED). Something like dok.get("fname") should do it. Erick On 6/1/07, starz10de <[EMAIL PROTECTED]> wrote: Hello all, I am printing luecene index content and I succes

question about lucene

Hi, I've just downloaded Lucene, tried demo and looked at the documentation. The Indexing and Searching work great and fast but I also need to display all the actual "hits": the lines from the files that match a particular query. Does Lucene provide means to do it? Thanks a lot, Tanya

Re: Lucene index content

Try Luke: http://www.getopt.org/luke/ It is a great tool to inspect the index contents. Luis. starz10de escribió: Hello all, I am printing luecene index content and I successed but I don't know how to print the indexed file names. System.out.println(dok.doc() ); here it printed the

Lucene index content

Hello all, I am printing luecene index content and I successed but I don't know how to print the indexed file names. System.out.println(dok.doc() ); here it printed the doc ID , but I need the document name. for exxample doc ID =1 , the file name = F1, how to print the file name F1. than

Re: exception during optimze

"Cedric Ho" <[EMAIL PROTECTED]> wrote: > When I tried to build an index last night, the following exception > occurred during call to IndexWriter.optimze(): > > java.lang.NullPointerException > at > org.apache.lucene.index.IndexFileDeleter.findDeletableFiles(IndexFileDeleter.java:88) >