Hi! I have this NotEmptyQuery class (http://gist.github.com/78115) which extends the MultiTermQuery. The class is added into a BooleanQuery, after some other queries (e.g. after TermQuery and LongTrieRangeFilter queries). I wonder: does Lucene need to scan all the terms in the inverted index and the collect all the document identifiers into the DocIdSet in order to implement the MultiTermQuery which goes after some other queries in a BooleanQuery? Like, collecting a thouthand of document identifiers only to filter a few documents which remain after the other queries had been fired up? Or does Lucene use some trickery to only provide a subset of terms to the MultiTermQuery?
Sorry if this is a dumb question or a part of a FAQ somewhere. (I haven't found any related performance discussion of Lucene internals on the site). --------------------------------------------------------------------- To unsubscribe, e-mail: java-user-unsubscr...@lucene.apache.org For additional commands, e-mail: java-user-h...@lucene.apache.org
