/* * 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.jackrabbit.core.query.lucene; import org.apache.jackrabbit.core.id.NodeId; import org.apache.jackrabbit.core.SessionImpl; import org.apache.jackrabbit.core.query.lucene.hits.AbstractHitCollector; import org.apache.jackrabbit.core.query.lucene.hits.AdaptingHits; import org.apache.jackrabbit.core.query.lucene.hits.Hits; import org.apache.jackrabbit.core.query.lucene.hits.ScorerHits; import org.apache.jackrabbit.core.state.ItemStateException; import org.apache.jackrabbit.core.state.ItemStateManager; import org.apache.jackrabbit.core.state.NodeState; import org.apache.jackrabbit.core.state.ChildNodeEntry; import org.apache.jackrabbit.spi.Name; import org.apache.jackrabbit.spi.commons.query.LocationStepQueryNode; import org.apache.lucene.document.Document; import org.apache.lucene.index.IndexReader; import org.apache.lucene.index.Term; import org.apache.lucene.index.TermDocs; import org.apache.lucene.search.Explanation; import org.apache.lucene.search.Query; import org.apache.lucene.search.Scorer; import org.apache.lucene.search.Searcher; import org.apache.lucene.search.Similarity; import org.apache.lucene.search.Weight; import org.apache.lucene.search.Sort; import org.slf4j.Logger; import org.slf4j.LoggerFactory; import java.io.IOException; import java.util.List; import java.util.Set; import java.util.HashMap; import java.util.Map; import java.util.HashSet; import java.util.ArrayList; /** * Implements a lucene <code>Query</code> which returns the child nodes of the * nodes selected by another <code>Query</code>. */ @SuppressWarnings("serial") class ChildAxisQuery extends Query implements JackrabbitQuery { /** * The logger instance for this class. */ private static final Logger log = LoggerFactory.getLogger(ChildAxisQuery.class); /** * Threshold when children calculation is switched to * {@link HierarchyResolvingChildrenCalculator}. */ private static int CONTEXT_SIZE_THRESHOLD = 10; /** * The item state manager containing persistent item states. */ private final ItemStateManager itemMgr; /** * The context query */ private Query contextQuery; /** * The nameTest to apply on the child axis, or <code>null</code> if all * child nodes should be selected. */ private final Name nameTest; /** * The context position for the selected child node, or * {@link LocationStepQueryNode#NONE} if no position is specified. */ private final int position; /** * The index format version. */ private final IndexFormatVersion version; /** * The internal namespace mappings. */ private final NamespaceMappings nsMappings; /** * The scorer of the context query */ private Scorer contextScorer; /** * The scorer of the name test query */ private Scorer nameTestScorer; /** * Creates a new <code>ChildAxisQuery</code> based on a <code>context</code> * query. * * @param itemMgr the item state manager. * @param context the context for this query. * @param nameTest a name test or <code>null</code> if any child node is * selected. * @param version the index format version. * @param nsMappings the internal namespace mappings. */ ChildAxisQuery(ItemStateManager itemMgr, Query context, Name nameTest, IndexFormatVersion version, NamespaceMappings nsMappings) { this(itemMgr, context, nameTest, LocationStepQueryNode.NONE, version, nsMappings); } /** * Creates a new <code>ChildAxisQuery</code> based on a <code>context</code> * query. * * @param itemMgr the item state manager. * @param context the context for this query. * @param nameTest a name test or <code>null</code> if any child node is * selected. * @param position the context position of the child node to select. If * <code>position</code> is {@link LocationStepQueryNode#NONE}, the context * position of the child node is not checked. * @param version the index format version. * @param nsMapping the internal namespace mappings. */ ChildAxisQuery(ItemStateManager itemMgr, Query context, Name nameTest, int position, IndexFormatVersion version, NamespaceMappings nsMapping) { this.itemMgr = itemMgr; this.contextQuery = context; this.nameTest = nameTest; this.position = position; this.version = version; this.nsMappings = nsMapping; } /** * @return the context query of this child axis query. */ Query getContextQuery() { return contextQuery; } /** * @return <code>true</code> if this child axis query matches any child * node; <code>false</code> otherwise. */ boolean matchesAnyChildNode() { return nameTest == null && position == LocationStepQueryNode.NONE; } /** * @return the name test or <code>null</code> if none was specified. */ Name getNameTest() { return nameTest; } /** * @return the position check or {@link LocationStepQueryNode#NONE} is none * was specified. */ int getPosition() { return position; } /** * Creates a <code>Weight</code> instance for this query. * * @param searcher the <code>Searcher</code> instance to use. * @return a <code>ChildAxisWeight</code>. */ public Weight createWeight(Searcher searcher) { return new ChildAxisWeight(searcher); } /** * {@inheritDoc} */ public void extractTerms(Set<Term> terms) { contextQuery.extractTerms(terms); } /** * {@inheritDoc} */ public Query rewrite(IndexReader reader) throws IOException { Query cQuery = contextQuery.rewrite(reader); // only try to compact if no position is specified if (position == LocationStepQueryNode.NONE) { if (cQuery instanceof DescendantSelfAxisQuery) { DescendantSelfAxisQuery dsaq = (DescendantSelfAxisQuery) cQuery; if (dsaq.subQueryMatchesAll()) { Query sub; if (nameTest == null) { sub = new MatchAllDocsQuery(); } else { sub = new NameQuery(nameTest, version, nsMappings); } return new DescendantSelfAxisQuery(dsaq.getContextQuery(), sub, dsaq.getMinLevels() + 1).rewrite(reader); } } } // if we get here we could not compact the query if (cQuery == contextQuery) { return this; } else { return new ChildAxisQuery(itemMgr, cQuery, nameTest, position, version, nsMappings); } } /** * {@inheritDoc} */ public String toString(String field) { StringBuffer sb = new StringBuffer(); sb.append("ChildAxisQuery("); sb.append(contextQuery); sb.append(", "); sb.append(nameTest); if (position != LocationStepQueryNode.NONE) { sb.append(", "); sb.append(position); } sb.append(")"); return sb.toString(); } //-------------------< JackrabbitQuery >------------------------------------ /** * {@inheritDoc} */ public QueryHits execute(JackrabbitIndexSearcher searcher, SessionImpl session, Sort sort) throws IOException { if (sort.getSort().length == 0 && matchesAnyChildNode()) { Query context = getContextQuery(); return new ChildNodesQueryHits(searcher.evaluate(context), session); } else { return null; } } //-------------------< ChildAxisWeight >------------------------------------ /** * The <code>Weight</code> implementation for this <code>ChildAxisQuery</code>. */ private class ChildAxisWeight extends Weight { /** * The searcher in use */ private final Searcher searcher; /** * Creates a new <code>ChildAxisWeight</code> instance using * <code>searcher</code>. * * @param searcher a <code>Searcher</code> instance. */ private ChildAxisWeight(Searcher searcher) { this.searcher = searcher; } /** * Returns this <code>ChildAxisQuery</code>. * * @return this <code>ChildAxisQuery</code>. */ public Query getQuery() { return ChildAxisQuery.this; } /** * {@inheritDoc} */ public float getValue() { return 1.0f; } /** * {@inheritDoc} */ public float sumOfSquaredWeights() throws IOException { return 1.0f; } /** * {@inheritDoc} */ public void normalize(float norm) { } /** * Creates a scorer for this <code>ChildAxisQuery</code>. * * @param reader a reader for accessing the index. * @return a <code>ChildAxisScorer</code>. * @throws IOException if an error occurs while reading from the index. */ @Override public Scorer scorer(IndexReader reader, boolean scoreDocsInOrder, boolean topScorer) throws IOException { contextScorer = contextQuery.weight(searcher).scorer(reader, scoreDocsInOrder, false); if (nameTest != null) { nameTestScorer = new NameQuery(nameTest, version, nsMappings).weight(searcher).scorer(reader, scoreDocsInOrder, false); } return new ChildAxisScorer(searcher.getSimilarity(), reader, (HierarchyResolver) reader); } /** * {@inheritDoc} */ public Explanation explain(IndexReader reader, int doc) throws IOException { return new Explanation(); } } //----------------------< ChildAxisScorer >--------------------------------- /** * Implements a <code>Scorer</code> for this <code>ChildAxisQuery</code>. */ private class ChildAxisScorer extends Scorer { /** * An <code>IndexReader</code> to access the index. */ private final IndexReader reader; /** * The <code>HierarchyResolver</code> of the index. */ private final HierarchyResolver hResolver; /** * The next document id to return */ private int nextDoc = -1; /** * A <code>Hits</code> instance containing all hits */ private Hits hits; /** * Creates a new <code>ChildAxisScorer</code>. * * @param similarity the <code>Similarity</code> instance to use. * @param reader for index access. * @param hResolver the hierarchy resolver of <code>reader</code>. */ protected ChildAxisScorer(Similarity similarity, IndexReader reader, HierarchyResolver hResolver) { super(similarity); this.reader = reader; this.hResolver = hResolver; } @Override public int nextDoc() throws IOException { if (nextDoc == NO_MORE_DOCS) { return nextDoc; } calculateChildren(); do { nextDoc = hits.next(); } while (nextDoc > -1 && !indexIsValid(nextDoc)); if (nextDoc < 0) { nextDoc = NO_MORE_DOCS; } return nextDoc; } @Override public int docID() { return nextDoc; } @Override public float score() throws IOException { return 1.0f; } @Override public int advance(int target) throws IOException { if (nextDoc == NO_MORE_DOCS) { return nextDoc; } // optimize in the case of an advance to finish. // see https://issues.apache.org/jira/browse/JCR-3091 if (target == NO_MORE_DOCS) { hits.skipTo(target); nextDoc = NO_MORE_DOCS; return nextDoc; } calculateChildren(); nextDoc = hits.skipTo(target); if (nextDoc < 0) { nextDoc = NO_MORE_DOCS; } while (nextDoc != NO_MORE_DOCS && !indexIsValid(nextDoc)) { nextDoc(); } return nextDoc; } private void calculateChildren() throws IOException { if (hits == null) { final ChildrenCalculator[] calc = new ChildrenCalculator[1]; if (nameTestScorer == null) { // always use simple in that case calc[0] = new SimpleChildrenCalculator(reader, hResolver); contextScorer.score(new AbstractHitCollector() { @Override protected void collect(int doc, float score) { calc[0].collectContextHit(doc); } }); } else { // start simple but switch once threshold is reached calc[0] = new SimpleChildrenCalculator(reader, hResolver); contextScorer.score(new AbstractHitCollector() { private List<Integer> docIds = new ArrayList<Integer>(); @Override protected void collect(int doc, float score) { calc[0].collectContextHit(doc); if (docIds != null) { docIds.add(doc); if (docIds.size() > CONTEXT_SIZE_THRESHOLD) { // switch calc[0] = new HierarchyResolvingChildrenCalculator( reader, hResolver); for (int docId : docIds) { calc[0].collectContextHit(docId); } // indicate that we switched docIds = null; } } } }); } hits = calc[0].getHits(); } } private boolean indexIsValid(int i) throws IOException { if (position != LocationStepQueryNode.NONE) { Document node = reader.document(i, FieldSelectors.UUID_AND_PARENT); NodeId parentId = NodeId.valueOf(node.get(FieldNames.PARENT)); NodeId id = NodeId.valueOf(node.get(FieldNames.UUID)); try { NodeState state = (NodeState) itemMgr.getItemState(parentId); if (nameTest == null) { // only select this node if it is the child at // specified position List<ChildNodeEntry> childNodes = state.getChildNodeEntries(); if (position == LocationStepQueryNode.LAST) { // only select last if (childNodes.size() == 0 || !(childNodes.get(childNodes.size() - 1)).getId().equals(id)) { return false; } } else { if (position < 1 || childNodes.size() < position || !(childNodes.get(position - 1)).getId().equals(id)) { return false; } } } else { // select the node when its index is equal to // specified position if (position == LocationStepQueryNode.LAST) { // only select last ChildNodeEntry entry = state.getChildNodeEntry(id); if (entry == null) { // no such child node, probably deleted meanwhile return false; } else { // only use the last one Name name = entry.getName(); List<ChildNodeEntry> childNodes = state.getChildNodeEntries(name); if (childNodes.size() == 0 || !(childNodes.get(childNodes.size() - 1)).getId().equals(id)) { return false; } } } else { ChildNodeEntry entry = state.getChildNodeEntry(id); if (entry == null) { // no such child node, probably has been deleted meanwhile return false; } else { if (entry.getIndex() != position) { return false; } } } } } catch (ItemStateException e) { // ignore this node, probably has been deleted meanwhile return false; } } return true; } } /** * Base class to calculate the children for a context query. */ private abstract class ChildrenCalculator { /** * The current index reader. */ protected final IndexReader reader; /** * The current hierarchy resolver. */ protected final HierarchyResolver hResolver; /** * Creates a new children calculator with the given index reader and * hierarchy resolver. * * @param reader the current index reader. * @param hResolver the current hierarchy resolver. */ public ChildrenCalculator(IndexReader reader, HierarchyResolver hResolver) { this.reader = reader; this.hResolver = hResolver; } /** * Collects a context hit. * * @param doc the lucene document number of the context hit. */ protected abstract void collectContextHit(int doc); /** * @return the hits that contains the children. * @throws IOException if an error occurs while reading from the index. */ public abstract Hits getHits() throws IOException; } /** * An implementation of a children calculator using the item state manager. */ private final class SimpleChildrenCalculator extends ChildrenCalculator { /** * The context hits. */ private final Hits contextHits = new AdaptingHits(); /** * Creates a new simple children calculator. * * @param reader the current index reader. * @param hResolver the current hierarchy resolver. */ public SimpleChildrenCalculator(IndexReader reader, HierarchyResolver hResolver) { super(reader, hResolver); } /** * {@inheritDoc} */ protected void collectContextHit(int doc) { contextHits.set(doc); } /** * {@inheritDoc} */ public Hits getHits() throws IOException { // read the uuids of the context nodes Map<Integer, String> uuids = new HashMap<Integer, String>(); for (int i = contextHits.next(); i > -1; i = contextHits.next()) { String uuid = reader.document(i, FieldSelectors.UUID).get(FieldNames.UUID); uuids.put(i, uuid); } // get child node entries for each hit Hits childrenHits = new AdaptingHits(); for (String uuid : uuids.values()) { NodeId id = new NodeId(uuid); try { long time = System.currentTimeMillis(); NodeState state = (NodeState) itemMgr.getItemState(id); time = System.currentTimeMillis() - time; log.debug("got NodeState with id {} in {} ms.", id, time); List<ChildNodeEntry> entries; if (nameTest != null) { entries = state.getChildNodeEntries(nameTest); } else { // get all children entries = state.getChildNodeEntries(); } for (ChildNodeEntry entry : entries) { NodeId childId = entry.getId(); Term uuidTerm = TermFactory.createUUIDTerm(childId.toString()); TermDocs docs = reader.termDocs(uuidTerm); try { if (docs.next()) { childrenHits.set(docs.doc()); } } finally { docs.close(); } } } catch (ItemStateException e) { // does not exist anymore -> ignore } } return childrenHits; } } /** * An implementation of a children calculator that uses the hierarchy * resolver. This implementation requires that * {@link ChildAxisQuery#nameTestScorer} is non null. */ private final class HierarchyResolvingChildrenCalculator extends ChildrenCalculator { /** * The document numbers of the context hits. */ private final Set<Integer> docIds = new HashSet<Integer>(); /** * Creates a new hierarchy resolving children calculator. * * @param reader the current index reader. * @param hResolver the current hierarchy resolver. */ public HierarchyResolvingChildrenCalculator(IndexReader reader, HierarchyResolver hResolver) { super(reader, hResolver); } /** * {@inheritDoc} */ protected void collectContextHit(int doc) { docIds.add(doc); } /** * {@inheritDoc} */ public Hits getHits() throws IOException { long time = System.currentTimeMillis(); Hits childrenHits = new AdaptingHits(); Hits nameHits = new ScorerHits(nameTestScorer); int[] docs = new int[1]; for (int h = nameHits.next(); h > -1; h = nameHits.next()) { docs = hResolver.getParents(h, docs); if (docs.length == 1) { // optimize single value if (docIds.contains(docs[0])) { childrenHits.set(h); } } else { for (int i = 0; i < docs.length; i++) { if (docIds.contains(docs[i])) { childrenHits.set(h); } } } } time = System.currentTimeMillis() - time; log.debug("Filtered hits in {} ms.", time); return childrenHits; } } }