cpoerschke commented on a change in pull request #300: SOLR-11831: Skip second
grouping step if group.limit is 1 (aka Las Vegas Patch)
URL: https://github.com/apache/lucene-solr/pull/300#discussion_r311202450
##########
File path:
solr/core/src/java/org/apache/solr/search/grouping/distributed/shardresultserializer/SearchGroupsResultTransformer.java
##########
@@ -142,4 +150,58 @@ private NamedList
serializeSearchGroup(Collection<SearchGroup<BytesRef>> data, S
return result;
}
+ public static class SkipSecondStepSearchResultResultTransformer extends
SearchGroupsResultTransformer {
+
+ private static final String TOP_DOC_SOLR_ID_KEY = "topDocSolrId";
+ private static final String TOP_DOC_SCORE_KEY = "topDocScore";
+ private static final String SORTVALUES_KEY = "sortValues";
+
+ private final SchemaField uniqueField;
+
+ public SkipSecondStepSearchResultResultTransformer(SolrIndexSearcher
searcher) {
+ super(searcher);
+ this.uniqueField = searcher.getSchema().getUniqueKeyField();
+ }
+
+ @Override
+ protected Object[] getSortValues(Object groupDocs) {
+ NamedList<Object> groupInfo = (NamedList) groupDocs;
+ final ArrayList<?> sortValues = (ArrayList<?>)
groupInfo.get(SORTVALUES_KEY);
+ return sortValues.toArray(new Comparable[sortValues.size()]);
+ }
+
+ @Override
+ protected SearchGroup<BytesRef> deserializeOneSearchGroup(SchemaField
groupField, String groupValue,
+ SortField[]
groupSortField, Object rawSearchGroupData) {
+ SearchGroup<BytesRef> searchGroup =
super.deserializeOneSearchGroup(groupField, groupValue, groupSortField,
rawSearchGroupData);
+ NamedList<Object> groupInfo = (NamedList) rawSearchGroupData;
+ searchGroup.topDocLuceneId = DocIdSetIterator.NO_MORE_DOCS;
+ searchGroup.topDocScore = (float) groupInfo.get(TOP_DOC_SCORE_KEY);
+ searchGroup.topDocSolrId = groupInfo.get(TOP_DOC_SOLR_ID_KEY);
+ return searchGroup;
+ }
+
+ @Override
+ protected Object serializeOneSearchGroup(SortField[] groupSortField,
SearchGroup<BytesRef> searchGroup) {
+ Document luceneDoc = null;
+ /** Use the lucene id to get the unique solr id so that it can be sent
to the federator.
+ * The lucene id of a document is not unique across all shards i.e.
different documents
+ * in different shards could have the same lucene id, whereas the solr
id is guaranteed
+ * to be unique so this is what we need to return to the federator
+ **/
+ try {
+ luceneDoc = searcher.doc(searchGroup.topDocLuceneId,
Collections.singleton(uniqueField.getName()));
Review comment:
https://github.com/cpoerschke/lucene-solr/commit/20129e7d3f7e12f442254e780e7da9a590a9036b
proposes to introduce a `uniqueFieldNameAsSet` member to avoid successive
calls allocating identical singleton sets.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
