gsmiller commented on code in PR #13568: URL: https://github.com/apache/lucene/pull/13568#discussion_r1693205018
########## lucene/sandbox/src/java/org/apache/lucene/sandbox/facet/recorders/FacetRecorder.java: ########## @@ -0,0 +1,55 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +package org.apache.lucene.sandbox.facet.recorders; + +import java.io.IOException; +import org.apache.lucene.index.LeafReaderContext; +import org.apache.lucene.sandbox.facet.cutters.LeafFacetCutter; +import org.apache.lucene.sandbox.facet.misc.FacetRollup; +import org.apache.lucene.sandbox.facet.ordinals.OrdinalIterator; + +/** + * Record data for each facet of each doc. + * + * <p>TODO: In the next iteration we can add an extra layer between FacetRecorder and + * LeafFacetRecorder, e.g. SliceFacetRecorder. The new layer will be created per {@link + * org.apache.lucene.search.Collector}, which means that collecting of multiple leafs (segments) + * within a slice is sequential and can be done to a single non-sync map to improve performance and + * reduce memory consumption. We already tried that, but didn't see any performance improvement. + * Given that it also makes lazy leaf recorder init in {@link + * org.apache.lucene.sandbox.facet.FacetFieldCollector} trickier, it was decided to rollback the + * initial attempt and try again later, in the next iteration. + */ +public interface FacetRecorder { + /** Get leaf recorder. */ + LeafFacetRecorder getLeafRecorder(LeafReaderContext context) throws IOException; + + /** Return next collected ordinal, or {@link LeafFacetCutter#NO_MORE_ORDS} */ + OrdinalIterator recordedOrds(); + + /** True if there are no records */ + boolean isEmpty(); Review Comment: Got it. I'm still not sure how useful `isEmpty` is though? It would only tell you that a recorder didn't record anything right? I wonder how common that is? As a counter point, a common faceting use-case with taxonomy faceting is to pack all dimensions into one index field. If a user is facing on a specific set of dimensions in a global field like this, it can also be the case that the recorders "sees" plenty of facets but none in the dimensions the user cares about. The `isEmpty` method doesn't cover that case at all (you still have to create an iterator over the dims you care about and then find out it's empty). So I'm just not sure how useful it is, and advocate for not adding speculative API methods unless there's a solid use-case as it can make future extension cumbersome/awkward. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org For additional commands, e-mail: issues-h...@lucene.apache.org