dsmiley commented on a change in pull request #1142: SOLR-14166: fq cache=false
should use TwoPhaseIterator
URL: https://github.com/apache/lucene-solr/pull/1142#discussion_r363293567
##########
File path:
solr/contrib/ltr/src/java/org/apache/solr/ltr/feature/SolrFeature.java
##########
@@ -237,21 +237,25 @@ public FeatureScorer scorer(LeafReaderContext context)
throws IOException {
* @return DocIdSetIterator to traverse documents that matched all filter
* criteria
*/
+ // TODO it's not optimal to call getProcessedFilter per-segment! Save the
results into one Query
+ // TODO rename to "FromFilterQueries" suffix to at least suggest this uses
the filter cache
private DocIdSetIterator getDocIdSetIteratorFromQueries(List<Query>
queries,
LeafReaderContext context) throws IOException {
final SolrIndexSearcher.ProcessedFilter pf = ((SolrIndexSearcher)
searcher)
.getProcessedFilter(null, queries);
- final Bits liveDocs = context.reader().getLiveDocs();
-
- DocIdSetIterator idIter = null;
- if (pf.filter != null) {
- final DocIdSet idSet = pf.filter.getDocIdSet(context, liveDocs);
- if (idSet != null) {
- idIter = idSet.iterator();
- }
+ if (pf.postFilter != null) {
+ throw new SolrException(SolrException.ErrorCode.BAD_REQUEST,
+ "PostFilter queries are not supported");
}
-
- return idIter;
+ Query q = pf.filter;
+ if (q == null) {
+ q = new MatchAllDocsQuery(); // usually never happens?
+ }
+ Scorer scorer = q.createWeight(searcher, ScoreMode.COMPLETE_NO_SCORES,
1f).scorer(context);
+ if (scorer != null) {
+ return scorer.iterator();
Review comment:
This particular method on this class for LTR wants to return a
DocIdSetIterator. You are correct that this method will not completely benefit
from TwoPhaseIterator as-designed. It will benefit from the cost ordering
aspect though. I like your suggestion of returning a Scorer, thus enabling the
caller to _potentially_ use it better (it does not today). But I don't want to
scope creep this PR into the LTR module more than necessary to accomplish the
primary goal of the PR. If what you propose is pretty simple then it can be
done now but I don't see it.
----------------------------------------------------------------
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:
us...@infra.apache.org
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org
For additional commands, e-mail: issues-h...@lucene.apache.org
