initial facet refactoring

the main goal of the facet refactoring is to allow for two modes of facet execution, collector based, that get callbacks as hist match, and post based, which iterates over all the relevant hits it also includes a some simplification of the facet implementation

Author: kimchy

src/main/java/org/elasticsearch/action/search/SearchRequestBuilder.java

@@ -33,7 +33,7 @@
 import org.elasticsearch.index.query.QueryBuilder;
 import org.elasticsearch.search.Scroll;
 import org.elasticsearch.search.builder.SearchSourceBuilder;
-import org.elasticsearch.search.facet.AbstractFacetBuilder;
+import org.elasticsearch.search.facet.FacetBuilder;
 import org.elasticsearch.search.highlight.HighlightBuilder;
 import org.elasticsearch.search.rescore.RescoreBuilder;
 import org.elasticsearch.search.sort.SortBuilder;
@@ -485,7 +485,7 @@ public SearchRequestBuilder addFields(String... fields) {
  /**
  * Adds a facet to the search operation.
  */
- public SearchRequestBuilder addFacet(AbstractFacetBuilder facet) {
+ public SearchRequestBuilder addFacet(FacetBuilder facet) {
  sourceBuilder().facet(facet);
  return this;
  }
@@ -663,17 +663,17 @@ public SearchRequestBuilder addSuggestion(SuggestBuilder.Suggestion suggestion)
  suggestBuilder().addSuggestion(suggestion);
  return this;
  }
- 
+
  public SearchRequestBuilder setRescorer(RescoreBuilder.Rescorer rescorer) {
  rescoreBuilder().setRescorer(rescorer);
  return this;
  }
- 
+
  public SearchRequestBuilder setRescoreWindow(int window) {
  rescoreBuilder().setWindowSize(window);
  return this;
  }
- 
+
  /**
  * Sets the source of the request as a json string. Note, settings anything other
  * than the search type will cause this source to be overridden, consider using
@@ -850,7 +850,7 @@ private HighlightBuilder highlightBuilder() {
  private SuggestBuilder suggestBuilder() {
  return sourceBuilder().suggest();
  }
- 
+
  private RescoreBuilder rescoreBuilder() {
  return sourceBuilder().rescore();
  }

src/main/java/org/elasticsearch/common/lucene/MultiCollector.java

@@ -23,13 +23,14 @@
 import org.apache.lucene.search.Collector;
 import org.apache.lucene.search.ScoreCachingWrappingScorer;
 import org.apache.lucene.search.Scorer;
+import org.elasticsearch.common.lucene.search.XCollector;
 
 import java.io.IOException;
 
 /**
  *
  */
-public class MultiCollector extends Collector {
+public class MultiCollector extends XCollector {
 
  private final Collector collector;
 
@@ -80,4 +81,16 @@ public boolean acceptsDocsOutOfOrder() {
  }
  return true;
  }
+
+ @Override
+ public void postCollection() {
+ if (collector instanceof XCollector) {
+ ((XCollector) collector).postCollection();
+ }
+ for (Collector collector : collectors) {
+ if (collector instanceof XCollector) {
+ ((XCollector) collector).postCollection();
+ }
+ }
+ }
 }

src/main/java/org/elasticsearch/search/facet/FacetCollector.java renamed to src/main/java/org/elasticsearch/common/lucene/docset/ContextDocIdSet.java

@@ -17,17 +17,20 @@
  * under the License.
  */
 
-package org.elasticsearch.search.facet;
+package org.elasticsearch.common.lucene.docset;
 
-import org.apache.lucene.search.Collector;
-import org.apache.lucene.search.Filter;
+import org.apache.lucene.index.AtomicReaderContext;
+import org.apache.lucene.search.DocIdSet;
 
 /**
- *
  */
-public abstract class FacetCollector extends Collector {
+public class ContextDocIdSet {
 
- public abstract Facet facet();
+ public final AtomicReaderContext context;
+ public final DocIdSet docSet;
 
- public abstract void setFilter(Filter filter);
+ public ContextDocIdSet(AtomicReaderContext context, DocIdSet docSet) {
+ this.context = context;
+ this.docSet = docSet;
+ }
 }

src/main/java/org/elasticsearch/common/lucene/search/FilteredCollector.java

@@ -31,7 +31,7 @@
 /**
  *
  */
-public class FilteredCollector extends Collector {
+public class FilteredCollector extends XCollector {
 
  private final Collector collector;
 
@@ -44,6 +44,13 @@ public FilteredCollector(Collector collector, Filter filter) {
  this.filter = filter;
  }
 
+ @Override
+ public void postCollection() {
+ if (collector instanceof XCollector) {
+ ((XCollector) collector).postCollection();
+ }
+ }
+
  @Override
  public void setScorer(Scorer scorer) throws IOException {
  collector.setScorer(scorer);

src/main/java/org/elasticsearch/search/facet/FacetProcessor.java renamed to src/main/java/org/elasticsearch/common/lucene/search/XCollector.java

@@ -17,19 +17,17 @@
  * under the License.
  */
 
-package org.elasticsearch.search.facet;
+package org.elasticsearch.common.lucene.search;
 
-import org.elasticsearch.common.xcontent.XContentParser;
-import org.elasticsearch.search.internal.SearchContext;
-
-import java.io.IOException;
+import org.apache.lucene.search.Collector;
 
 /**
- *
+ * An extension to {@link Collector} that allows for a callback when
+ * collection is done.
  */
-public interface FacetProcessor {
+public abstract class XCollector extends Collector {
 
- String[] types();
+ public void postCollection() {
 
- FacetCollector parse(String facetName, XContentParser parser, SearchContext context) throws IOException;
+ }
 }

src/main/java/org/elasticsearch/index/cache/IndexCache.java

@@ -31,6 +31,7 @@
 import org.elasticsearch.common.unit.TimeValue;
 import org.elasticsearch.index.AbstractIndexComponent;
 import org.elasticsearch.index.Index;
+import org.elasticsearch.index.cache.docset.DocSetCache;
 import org.elasticsearch.index.cache.filter.FilterCache;
 import org.elasticsearch.index.cache.id.IdCache;
 import org.elasticsearch.index.cache.query.parser.QueryParserCache;
@@ -42,24 +43,24 @@
 public class IndexCache extends AbstractIndexComponent implements CloseableComponent, ClusterStateListener {
 
  private final FilterCache filterCache;
-
  private final QueryParserCache queryParserCache;
-
  private final IdCache idCache;
+ private final DocSetCache docSetCache;
 
  private final TimeValue refreshInterval;
-
  private ClusterService clusterService;
 
  private long latestCacheStatsTimestamp = -1;
  private CacheStats latestCacheStats;
 
  @Inject
- public IndexCache(Index index, @IndexSettings Settings indexSettings, FilterCache filterCache, QueryParserCache queryParserCache, IdCache idCache) {
+ public IndexCache(Index index, @IndexSettings Settings indexSettings, FilterCache filterCache, QueryParserCache queryParserCache, IdCache idCache,
+ DocSetCache docSetCache) {
  super(index, indexSettings);
  this.filterCache = filterCache;
  this.queryParserCache = queryParserCache;
  this.idCache = idCache;
+ this.docSetCache = docSetCache;
 
  this.refreshInterval = componentSettings.getAsTime("stats.refresh_interval", TimeValue.timeValueSeconds(1));
 
@@ -94,6 +95,10 @@ public FilterCache filter() {
  return filterCache;
  }
 
+ public DocSetCache docSet() {
+ return this.docSetCache;
+ }
+
  public IdCache idCache() {
  return this.idCache;
  }
@@ -107,6 +112,7 @@ public void close() throws ElasticSearchException {
  filterCache.close();
  idCache.close();
  queryParserCache.close();
+ docSetCache.clear("close");
  if (clusterService != null) {
  clusterService.remove(this);
  }
@@ -115,12 +121,14 @@ public void close() throws ElasticSearchException {
  public void clear(IndexReader reader) {
  filterCache.clear(reader);
  idCache.clear(reader);
+ docSetCache.clear(reader);
  }
 
  public void clear(String reason) {
  filterCache.clear(reason);
  idCache.clear();
  queryParserCache.clear();
+ docSetCache.clear(reason);
  }
 
  @Override

src/main/java/org/elasticsearch/index/cache/IndexCacheModule.java

@@ -21,6 +21,7 @@
 
 import org.elasticsearch.common.inject.AbstractModule;
 import org.elasticsearch.common.settings.Settings;
+import org.elasticsearch.index.cache.docset.DocSetCacheModule;
 import org.elasticsearch.index.cache.filter.FilterCacheModule;
 import org.elasticsearch.index.cache.id.IdCacheModule;
 import org.elasticsearch.index.cache.query.parser.QueryParserCacheModule;
@@ -41,6 +42,7 @@ protected void configure() {
  new FilterCacheModule(settings).configure(binder());
  new IdCacheModule(settings).configure(binder());
  new QueryParserCacheModule(settings).configure(binder());
+ new DocSetCacheModule(settings).configure(binder());
 
  bind(IndexCache.class).asEagerSingleton();
  }

src/main/java/org/elasticsearch/search/facet/OptimizeGlobalFacetCollector.java renamed to src/main/java/org/elasticsearch/index/cache/docset/DocSetCache.java

@@ -17,13 +17,22 @@
  * under the License.
  */
 
-package org.elasticsearch.search.facet;
+package org.elasticsearch.index.cache.docset;
 
-import org.elasticsearch.search.internal.SearchContext;
+import org.apache.lucene.index.AtomicReaderContext;
+import org.apache.lucene.index.IndexReader;
+import org.elasticsearch.common.lucene.docset.ContextDocIdSet;
+import org.elasticsearch.index.IndexComponent;
 
-import java.io.IOException;
+/**
+ */
+public interface DocSetCache extends IndexComponent {
+
+ void clear(String reason);
+
+ void clear(IndexReader reader);
 
-public interface OptimizeGlobalFacetCollector {
+ ContextDocIdSet obtain(AtomicReaderContext context);
 
- void optimizedGlobalExecution(SearchContext searchContext) throws IOException;
-}
+ void release(ContextDocIdSet docSet);
+}

src/main/java/org/elasticsearch/index/cache/docset/DocSetCacheModule.java

@@ -0,0 +1,44 @@
+/*
+ * Licensed to ElasticSearch and Shay Banon under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. ElasticSearch 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.elasticsearch.index.cache.docset;
+
+import org.elasticsearch.common.inject.AbstractModule;
+import org.elasticsearch.common.inject.Scopes;
+import org.elasticsearch.common.settings.Settings;
+import org.elasticsearch.index.cache.docset.simple.SimpleDocSetCache;
+
+/**
+ *
+ */
+public class DocSetCacheModule extends AbstractModule {
+
+ private final Settings settings;
+
+ public DocSetCacheModule(Settings settings) {
+ this.settings = settings;
+ }
+
+ @Override
+ protected void configure() {
+ bind(DocSetCache.class)
+ .to(settings.getAsClass("index.cache.docset.type", SimpleDocSetCache.class, "org.elasticsearch.index.cache.docset.", "DocSetCache"))
+ .in(Scopes.SINGLETON);
+ }
+}

src/main/java/org/elasticsearch/index/cache/docset/none/NoneDocSetCache.java

@@ -0,0 +1,58 @@
+/*
+ * Licensed to ElasticSearch and Shay Banon under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. ElasticSearch 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.elasticsearch.index.cache.docset.none;
+
+import org.apache.lucene.index.AtomicReaderContext;
+import org.apache.lucene.index.IndexReader;
+import org.apache.lucene.util.FixedBitSet;
+import org.elasticsearch.common.inject.Inject;
+import org.elasticsearch.common.lucene.docset.ContextDocIdSet;
+import org.elasticsearch.common.settings.Settings;
+import org.elasticsearch.index.AbstractIndexComponent;
+import org.elasticsearch.index.Index;
+import org.elasticsearch.index.cache.docset.DocSetCache;
+import org.elasticsearch.index.settings.IndexSettings;
+
+/**
+ */
+public class NoneDocSetCache extends AbstractIndexComponent implements DocSetCache {
+
+ @Inject
+ public NoneDocSetCache(Index index, @IndexSettings Settings indexSettings) {
+ super(index, indexSettings);
+ }
+
+ @Override
+ public void clear(String reason) {
+ }
+
+ @Override
+ public void clear(IndexReader reader) {
+ }
+
+ @Override
+ public ContextDocIdSet obtain(AtomicReaderContext context) {
+ return new ContextDocIdSet(context, new FixedBitSet(context.reader().maxDoc()));
+ }
+
+ @Override
+ public void release(ContextDocIdSet docSet) {
+ }
+}