You're calling .build once per suggestion, which is not right. Instead you should call it once, overall, and pass it an iterator that iterates over all the suggestions you pull from the index.
E.g. fill in TermFreqPayload[] up front by walking through your entire index, then create TermFreqPayloadArrayIterator from that and pass it to the .build method. Mike McCandless http://blog.mikemccandless.com On Sun, Jul 28, 2013 at 1:59 PM, vonPuh fonPuhendorf <vonpuhfonpuhend...@gmail.com> wrote: > Yes with the lookup api. It return the token with <b> </b> appended that`s > why it has to interact with api. But how to how to iterate over my real > index > In scala: > > def infixSuggest(){ > > val sourceindex = new File("/tmp/lucene/1374960475771") > val reader = DirectoryReader.open(FSDirectory.open(sourceindex)) > val searcher = new IndexSearcher(reader) > val suggester = new AnalyzingInfixSuggester(Version.LUCENE_44,new > File("/tmp/lucene/1374619766134"),analyzer,analyzer,3) > > val indReader = searcher.getIndexReader > val maxid = indReader.maxDoc() > > for(id<-0 until maxid){ > id+1 > > val doc = searcher.doc(id) > val keys = Array[TermFreqPayload](new > TermFreqPayload(doc.get("contents"), 3, new BytesRef(doc.get("id")))) > val iterator = new TermFreqPayloadArrayIterator(keys) > suggester.build(iterator) > > } > > } > > Figure out that AnalyzingInfixSuggester has to be in its own index and use > source index(the one i like to get words from it). It works fine but for > every increment its start from clean index i.e it tokenized and indexed > only the last document. Mike said that it has no incremental support, so > how to pass my source index. --------------------------------------------------------------------- To unsubscribe, e-mail: java-user-unsubscr...@lucene.apache.org For additional commands, e-mail: java-user-h...@lucene.apache.org
