Simplest solution is to wrap your findFeatures.reader in a SlowMultiReaderWrapper (as the exception suggests).
More performant solution is to change your code to visit the sequential sub-readers of findFeatures.reader, directly. But if performance isn't important here, just do the simple solution. Mike http://blog.mikemccandless.com On Wed, Mar 23, 2011 at 4:49 PM, nitinhardeniya <nitinharden...@gmail.com> wrote: > I have changed the code according to MIGRATE.txt > > but now i am getting an error at > > public long getCorpCount(Vector <SpanTermQuery> clauses) > { > long count=0; > try { > SpanQuery [] clause= new SpanQuery[clauses.size()]; > clause= clauses.toArray(clause); > // SpanNearQuery sq = new SpanNearQuery(clause,1,true); > SpanOrQuery sq=new > SpanOrQuery(clause); error here > Spans spans=sq.getSpans(findFeatures.reader); > > while(spans.next()) //here was the error > { > count++; > } > } > catch (Exception e) { > // TODO: handle exception > e.printStackTrace(); > } > > return count ; > } > > > the error log is : > > please use MultiFields.getDeletedDocs, or wrap your IndexReader with > SlowMultiReaderWrapper, if you really need a top level Bits deletedDocs > at > org.apache.lucene.index.DirectoryReader.getDeletedDocs(DirectoryReader.java:371) > at > org.apache.lucene.search.spans.SpanTermQuery.getSpans(SpanTermQuery.java:84) > at > org.apache.lucene.search.spans.SpanOrQuery$1.initSpanQueue(SpanOrQuery.java:172) > at > org.apache.lucene.search.spans.SpanOrQuery$1.next(SpanOrQuery.java:184) > at rankPhrase.calFeatures.Document.getCorpCount(Document.java:489) > at rankPhrase.calFeatures.Document.calculateTfCorpus(Document.java:468) > at > rankPhrase.calFeatures.findFeatures.computeAllFeatures(findFeatures.java:309) > at > rankPhrase.calFeatures.findFeatures.LoadPhrases(findFeatures.java:235) > at rankPhrase.calFeatures.findFeatures.main(findFeatures.java:380) > > On Wed, Mar 23, 2011 at 11:56 PM, Burton-West, Tom [via Lucene] < > ml-node+2721619-1942655904-77...@n3.nabble.com> wrote: > >> Hi, >> >> If I understand correctly what you are trying to do as far as getting >> corpusTF, you might want to look at the implementation of the "-t" flag in >> org.apache.lucene.misc/HighFreqTerms.java in contib. >> >> Take a look at the getTotalTermFreq method in trunk. >> >> >> >> http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/contrib/misc/src/java/org/apache/lucene/misc/HighFreqTerms.java?view=markup >> >> 3.x version here: >> >> >> http://svn.apache.org/viewvc/lucene/dev/branches/branch_3x/lucene/contrib/misc/src/java/org/apache/lucene/misc/HighFreqTerms.java?view=markup >> >> Tom >> http://www.hathitrust.org/blogs/large-scale-search >> >> >> >> -----Original Message----- >> From: nitinhardeniya [mailto:[hidden >> email]<http://user/SendEmail.jtp?type=node&node=2721619&i=0&by-user=t>] >> >> Sent: Tuesday, March 22, 2011 1:57 PM >> To: [hidden >> email]<http://user/SendEmail.jtp?type=node&node=2721619&i=1&by-user=t> >> Subject: TermDoc to TermDocsEnum >> >> hi >> >> I have a code that work fine with lucene 3.2 where i used TermDocs to find >> the corpusTF here is the code >> >> >> public void calculateCorpusTF(IndexReader reader) throws IOException { >> // TODO Auto-generated method stub >> Iterator it = word.iterator(); >> >> Iterator iwp = word_prop.iterator(); >> wordProp wp; >> Term ta = null; >> >> TermDocs tds; >> // DocsEnum tds; >> String text; >> tfDoc tfcoll; >> long freq=0; >> OpenBitSet skipDocs = null; >> skipDocs = new OpenBitSet(0); >> //System.out.println("Length: "+skipDocs.length()); >> try { >> >> while(it.hasNext()) >> { >> text=it.next(); >> wp=iwp.next(); >> >> System.out.println("Word is "+text); >> ta= new Term("content",text); >> //BytesRef term = new >> BytesRef(text.toCharArray(),0,text.length()); >> >> tfcoll = new tfDoc(); >> freq=0; >> >> >> tds=reader.termDocs(ta); >> // >> tds=reader.terms("content"); >> >> if(tds!=null) >> { >> >> while(tds.next()) >> { >> >> >> freq+=tds.freq(); >> // >> System.out.print( text +" "+ freq); >> } >> } >> >> // New Code --> >> >> // tds = reader.termDocsEnum(skipDocs, "content", term); >> // if(tds!=null) >> // { >> // while(true) { >> // freq += tds.freq(); >> // final int docID = tds.nextDoc(); >> // if (docID == DocsEnum.NO_MORE_DOCS) { >> // break; >> // } >> // } >> // } >> // >> // New code Ends <-- >> >> tfcoll.tfA=freq; >> System.out.print( text +" "+ freq); >> if(tfcoll.totalTF()==0) >> { >> //System.out.println(" >> "+tfcoll.tfA+" "+tfcoll.tfD+" "+tfcoll.tfC); >> System.out.println("Text "+text+ " >> Freq "+freq); >> } >> >> wp.tfColl=tfcoll; >> } >> } >> catch (Exception e) { >> // TODO: handle exception >> e.printStackTrace(); >> } >> >> } >> >> but now i have to use TermDocEnum because i am using lucene dev4.0 which >> does not have TermDocs method i was trying to change my code .please refer >> to new code [commented ] and tell me how to use this method in a proper way >> >> . if you can provide an example that would be great. >> tds = reader.termDocsEnum(skipDocs, "content", term); >> I have tried using null at skipdoc because i don't want to skip anything >> but >> it through error >> >> please help >> >> -- >> View this message in context: >> http://lucene.472066.n3.nabble.com/TermDoc-to-TermDocsEnum-tp2716046p2716046.html<http://lucene.472066.n3.nabble.com/TermDoc-to-TermDocsEnum-tp2716046p2716046.html?by-user=t> >> Sent from the Lucene - Java Users mailing list archive at Nabble.com. >> >> --------------------------------------------------------------------- >> To unsubscribe, e-mail: [hidden >> email]<http://user/SendEmail.jtp?type=node&node=2721619&i=2&by-user=t> >> For additional commands, e-mail: [hidden >> email]<http://user/SendEmail.jtp?type=node&node=2721619&i=3&by-user=t> >> >> >> --------------------------------------------------------------------- >> To unsubscribe, e-mail: [hidden >> email]<http://user/SendEmail.jtp?type=node&node=2721619&i=4&by-user=t> >> For additional commands, e-mail: [hidden >> email]<http://user/SendEmail.jtp?type=node&node=2721619&i=5&by-user=t> >> >> >> >> ------------------------------ >> If you reply to this email, your message will be added to the discussion >> below: >> >> http://lucene.472066.n3.nabble.com/TermDoc-to-TermDocsEnum-tp2716046p2721619.html >> To unsubscribe from TermDoc to TermDocsEnum, click >> here<http://lucene.472066.n3.nabble.com/template/NamlServlet.jtp?macro=unsubscribe_by_code&node=2716046&code=bml0aW5oYXJkZW5peWFAZ21haWwuY29tfDI3MTYwNDZ8LTkwNzg0MTk0MA==>. >> >> > > > > -- > Nitin Kumar Hardeniya > > M.Tech Computational Linguistics > IIIT Hyderabad > > SAVE PAPER - THINK BEFORE YOU PRINT > > > -- > View this message in context: > http://lucene.472066.n3.nabble.com/TermDoc-to-TermDocsEnum-tp2716046p2722185.html > Sent from the Lucene - Java Users mailing list archive at Nabble.com. --------------------------------------------------------------------- To unsubscribe, e-mail: java-user-unsubscr...@lucene.apache.org For additional commands, e-mail: java-user-h...@lucene.apache.org
