Aron, I believe we now have class ExtendedFieldCacheImpl extends FieldCacheImpl implements ExtendedFieldCache And this should support sorting by longs.
Otis -- Sematext -- http://sematext.com/ -- Lucene - Solr - Nutch ----- Original Message ---- From: Aron Sogor <[EMAIL PROTECTED]> To: java-user@lucene.apache.org Sent: Tuesday, January 15, 2008 8:37:49 AM Subject: Re: Sort does not work for me Lucky guy who gets the same problem. Found the issue: http://issues.apache.org/jira/browse/LUCENE-463 Lucene see numbrs in the field and thinks it is an int... than overflows the int. Force the sort field to be a SortField.String. Aron Sogor wrote: > Let me qualify my question: > > Sort is not working for a field that I stored : > > document.add(new Field(FIELD_RECEIVED, DateTools.timeToString( > System.currentTimeMillis(), > DateTools.Resolution.SECOND), > Field.Store.NO, Field.Index.UN_TOKENIZED)); > > using the query bellow with sort: > Sort sort = new Sort(FIELD_RECEIVED,true); > > even if I do reverse=false the order of the hits does not really > change. The question is why would that be. The dates are numbers that > even is sorted as strings should order the same, and reverse should work. > > Aron > > > > Aron Sogor wrote: >> Here is the document: >> >> MimeMessage mime = new MimeMessage(null, new >> FileInputStream(item >> .getMailFile())); >> document.add(new Field(FIELD_MAILID, >> item.getMailId().toString(), >> Field.Store.YES, Field.Index.UN_TOKENIZED)); >> document.add(new Field(FIELD_SUBJECT, mime.getSubject(), >> Field.Store.COMPRESS, Field.Index.TOKENIZED)); >> document.add(new Field(FIELD_FROM, >> getAddressLine(mime.getFrom()), >> Field.Store.COMPRESS, Field.Index.TOKENIZED)); >> document.add(new Field(FIELD_BODY, getContentReader(mime >> .getContent()))); >> document.add(new Field(FIELD_RECEIVED, >> DateTools.timeToString( >> System.currentTimeMillis(), >> DateTools.Resolution.SECOND), >> Field.Store.NO, Field.Index.UN_TOKENIZED)); >> document.add(new Field(FIELD_TAG, TAG_INBOX, Field.Store.YES, >> Field.Index.TOKENIZED)); >> return document; >> >> here is the search: >> >> public static Hits createInboxQuery(IndexSearcher searcher) >> throws ParseException, IOException { >> Query query = new QueryParser(FIELD_TAG, new StandardAnalyzer()) >> .parse(TAG_INBOX); >> Sort sort = new Sort(FIELD_RECEIVED,true); >> return searcher.search(query, sort); >> } >> >> IndexSearcher searcher = >> storage.getIndexSearcherForMailBox("junk"); >> Hits hits = >> DocumentAndQueryFactory.createInboxQuery(searcher); >> for (int i = 0; i < hits.length(); i++) { >> Document hit = (Document) hits.doc(i); >> >> System.out.println(hit.getField("subject").stringValue()); >> >> >> No matter what I do when I lit the hits they come back (these are the >> subjects/ but essentially same timestamp): >> >> Test MSGSun Jan 13 15:52:47 PST 2008 >> Test MSGSun Jan 13 15:52:54 PST 2008 >> Test MSGSun Jan 13 15:53:03 PST 2008 >> Test MSGSun Jan 13 15:53:12 PST 2008 >> Test MSGSun Jan 13 15:53:48 PST 2008 >> >> Setting the "revert" for the search Sort essentially has no effect. I >> perhaps very much missing something... but what? >> >> Aron >> >> --------------------------------------------------------------------- >> 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
