Github user dsmiley commented on a diff in the pull request:
https://github.com/apache/lucene-solr/pull/416#discussion_r204765898
--- Diff:
solr/core/src/java/org/apache/solr/response/transform/DeeplyNestedChildDocTransformer.java
---
@@ -0,0 +1,214 @@
+/*
+ * 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.solr.response.transform;
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.List;
+import java.util.Set;
+import java.util.stream.Collectors;
+
+import com.google.common.collect.ArrayListMultimap;
+import com.google.common.collect.Multimap;
+import org.apache.lucene.index.DocValues;
+import org.apache.lucene.index.IndexableField;
+import org.apache.lucene.index.LeafReaderContext;
+import org.apache.lucene.index.SortedDocValues;
+import org.apache.lucene.search.Query;
+import org.apache.lucene.search.Sort;
+import org.apache.lucene.search.SortField;
+import org.apache.lucene.search.join.BitSetProducer;
+import org.apache.lucene.search.join.ToChildBlockJoinQuery;
+import org.apache.lucene.util.BytesRef;
+import org.apache.solr.common.SolrDocument;
+import org.apache.solr.request.SolrQueryRequest;
+import org.apache.solr.response.DocsStreamer;
+import org.apache.solr.schema.FieldType;
+import org.apache.solr.schema.IndexSchema;
+import org.apache.solr.schema.SchemaField;
+import org.apache.solr.search.DocIterator;
+import org.apache.solr.search.DocList;
+import org.apache.solr.search.SolrDocumentFetcher;
+import org.apache.solr.search.SolrIndexSearcher;
+import org.apache.solr.search.SolrReturnFields;
+
+import static
org.apache.solr.response.transform.ChildDocTransformerFactory.NUM_SEP_CHAR;
+import static
org.apache.solr.response.transform.ChildDocTransformerFactory.PATH_SEP_CHAR;
+import static org.apache.solr.schema.IndexSchema.NEST_PATH_FIELD_NAME;
+
+class DeeplyNestedChildDocTransformer extends DocTransformer {
+
+ private final String name;
+ protected final SchemaField idField;
+ protected final SolrQueryRequest req;
+ protected final IndexSchema schema;
+ private BitSetProducer parentsFilter;
+ protected int limit;
+ private final static Sort docKeySort = new Sort(new SortField(null,
SortField.Type.DOC, false));
+ private Query childFilterQuery;
+
+ public DeeplyNestedChildDocTransformer(String name, final BitSetProducer
parentsFilter,
+ final SolrQueryRequest req, final
Query childFilterQuery, int limit) {
+ this.name = name;
+ this.schema = req.getSchema();
+ this.idField = this.schema.getUniqueKeyField();
+ this.req = req;
+ this.parentsFilter = parentsFilter;
+ this.limit = limit;
+ this.childFilterQuery = childFilterQuery;
+ }
+
+ @Override
+ public String getName() {
+ return name;
+ }
+
+ @Override
+ public String[] getExtraRequestFields() {
+ // we always need the idField (of the parent) in order to fill out
it's children
+ return new String[] { idField.getName() };
+ }
+
+ protected static String getSolrFieldString(Object fieldVal, FieldType
fieldType) {
+ return fieldVal instanceof IndexableField
+ ? fieldType.toExternal((IndexableField)fieldVal)
+ : fieldVal.toString();
+ }
+
+ @Override
+ public void transform(SolrDocument rootDoc, int rootDocId) {
+
+ FieldType idFt = idField.getType();
+
+ String rootIdExt =
getSolrFieldString(rootDoc.getFirstValue(idField.getName()), idFt);
+
+ try {
+ Query parentQuery = idFt.getFieldQuery(null, idField, rootIdExt);
+ Query query = new ToChildBlockJoinQuery(parentQuery, parentsFilter);
+ SolrIndexSearcher searcher = context.getSearcher();
+ DocList children = searcher.getDocList(query, childFilterQuery,
docKeySort, 0, limit);
+ long segAndId = searcher.lookupId(new BytesRef(rootIdExt));
+ final int seg = (int) (segAndId >> 32);
+ final LeafReaderContext leafReaderContext =
searcher.getIndexReader().leaves().get(seg);
+ final SortedDocValues segPathDocValues =
DocValues.getSorted(leafReaderContext.reader(), NEST_PATH_FIELD_NAME);
+
+ Multimap<String,SolrDocument> pendingParentPathsToChildren =
ArrayListMultimap.create();
+
+ if(children.matches() > 0) {
+ SolrDocumentFetcher docFetcher = searcher.getDocFetcher();
+ Set<String> dvFieldsToReturn =
docFetcher.getNonStoredDVs(true).stream()
+ .filter(name ->
!NEST_PATH_FIELD_NAME.equals(name)).collect(Collectors.toSet());
+ boolean shouldDecorateWithDVs = dvFieldsToReturn.size() > 0;
+ DocIterator i = children.iterator();
+ final int segBaseId = leafReaderContext.docBase;
+ final int firstChildDocId = i.nextDoc();
+ assert firstChildDocId < rootDocId;
+
+ for (int docId = firstChildDocId; docId < rootDocId; ++docId) {
+ // get the path
+ String fullDocPath = getPathByDocId(docId - segBaseId,
segPathDocValues);
+
+ // Is this doc a direct ancestor of another doc we've seen?
+ boolean isAncestor =
pendingParentPathsToChildren.containsKey(fullDocPath);
+
+ // Do we need to do anything with this doc (either ancestor or a
matched the child query)
+ if (isAncestor || children.exists(docId)) {
+ // load the doc
+ SolrDocument doc =
DocsStreamer.convertLuceneDocToSolrDoc(docFetcher.doc(docId),
+ schema, new SolrReturnFields());
+ doc.setField(NEST_PATH_FIELD_NAME, fullDocPath);
+ if (shouldDecorateWithDVs) {
+ docFetcher.decorateDocValueFields(doc, docId,
dvFieldsToReturn);
+ }
+ // get parent path
+ // put into pending
+ String parentDocPath = lookupParentPath(fullDocPath);
+ pendingParentPathsToChildren.put(parentDocPath, doc); //
multimap add (won't replace)
+
+ // if this path has pending child docs, add them.
+ if (isAncestor) {
+ addChildrenToParent(doc,
pendingParentPathsToChildren.get(fullDocPath));
+ pendingParentPathsToChildren.removeAll(fullDocPath); // no
longer pending
+ }
+ }
+ }
+
+ // only children of parent remain
+ assert pendingParentPathsToChildren.keySet().size() == 1;
+
+ addChildrenToParent(rootDoc,
pendingParentPathsToChildren.get(null));
+ }
+ } catch (IOException e) {
+ rootDoc.put(getName(), "Could not fetch child Documents");
+ }
+ }
+
+ void addChildToParent(SolrDocument parent, SolrDocument child, String
label) {
+ // lookup leaf key for these children using path
+ // depending on the label, add to the parent at the right key/label
+ // TODO: unfortunately this is the 2nd time we grab the paths for
these docs. resolve how?
+ String trimmedPath = trimSuffixFromPaths(getLastPath(label));
+ if (!parent.containsKey(trimmedPath) && (label.contains(NUM_SEP_CHAR)
&& !label.endsWith(NUM_SEP_CHAR))) {
+ List<SolrDocument> list = new ArrayList<>();
+ parent.setField(trimmedPath, list);
+ }
+ parent.addField(trimmedPath, child);
+ }
+
+ void addChildToParent(SolrDocument parent, SolrDocument child) {
+ String docPath =
getSolrFieldString(child.getFirstValue(NEST_PATH_FIELD_NAME),
schema.getFieldType(NEST_PATH_FIELD_NAME));
--- End diff --
I don't think we should assume that materialized SolrDocument contains any
particular field. The nest fields are internal details and furthermore one day
this transformer will likely have an "fl". So this means you should use
DocValues to fetch this field.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org
