I'm not sure that I understand the question. Can you not use a searcher based on the reader returned by IndexWriter.getReader() to determine if the doc is already in the index?
Or just use IndexWriter.updateDocument to save or replace as appropriate. -- Ian. On Tue, Feb 2, 2010 at 12:21 PM, Hayri <volkana...@gmail.com> wrote: > Ian Lea wrote: >> >> Sounds like a job for near realtime search aka NRT. >> >> Take a look at IndexWriter.getReader(). >> >> http://wiki.apache.org/lucene-java/NearRealtimeSearch >> >> http://www.lucidimagination.com/blog/2009/04/10/real-time-search-with-lucene/ >> >> And more with the help of your favourite search engine. >> >> >> -- >> Ian. >> >> On Mon, Feb 1, 2010 at 1:17 PM, Hayri <volkana...@gmail.com> wrote: >> >>> >>> Hi, >>> >>> I want to search an index and at the same time continue to my indexing. >>> ParallelReader doesn't solve my problem. >>> It is obvious that I am not searching multiple indexes at the same time. >>> How >>> can I build a document based lock, more over >>> I don't want to open and close and index every time while writing. >>> Merging >>> new index with the old one seems really time >>> consuming and also I want to search the new index. >>> >>> Is it possible? Please give hints... >>> >>> --------------------------------------------------------------------- >>> To unsubscribe, e-mail: java-user-unsubscr...@lucene.apache.org >>> For additional commands, e-mail: java-user-h...@lucene.apache.org >>> >>> >>> >> >> --------------------------------------------------------------------- >> To unsubscribe, e-mail: java-user-unsubscr...@lucene.apache.org >> For additional commands, e-mail: java-user-h...@lucene.apache.org >> >> >> > > Hi Thanks, > > I have another related question. I want to decide whether I can index the > document, > so basically when a new document comes I need to make a search in a specific > field in index. > The problem is I don't need to create a IndexSearcher to search for every > document to find if it indexed or not. > I can gather and index the documents to RAM for example. But after certain # > of documents ?? I have to decide to > merge the index except the documents that are already in file index (for > specific fields like unique). --------------------------------------------------------------------- To unsubscribe, e-mail: java-user-unsubscr...@lucene.apache.org For additional commands, e-mail: java-user-h...@lucene.apache.org
