Separate LeafCollector

This commit is contained in:
Andrea Cavalli 2021-09-25 14:09:10 +02:00
parent 5443e330bb
commit f03f7296d4
2 changed files with 36 additions and 28 deletions

View File

@ -28,7 +28,7 @@ public class AdaptiveLuceneMultiSearcher implements LuceneMultiSearcher {
} else if (queryParams.isSorted() || queryParams.isScored()) {
return scoredSimpleLuceneShardSearcher.collectMulti(indexSearchersMono, queryParams, keyFieldName, transformer);
} else {
if (queryParams.offset() + queryParams.limit() <= queryParams.pageLimits().getPageLimit(0)) {
if (((long) queryParams.offset() + (long) queryParams.limit()) <= (long) queryParams.pageLimits().getPageLimit(0)) {
// Run single-page searches using the paged multi searcher
return unsortedUnscoredPagedLuceneMultiSearcher.collectMulti(indexSearchersMono, queryParams, keyFieldName, transformer);
} else {

View File

@ -21,6 +21,8 @@ import org.apache.lucene.index.LeafReaderContext;
import org.apache.lucene.search.Collector;
import org.apache.lucene.search.CollectorManager;
import org.apache.lucene.search.IndexSearcher;
import org.apache.lucene.search.LeafCollector;
import org.apache.lucene.search.Scorable;
import org.apache.lucene.search.ScoreDoc;
import org.apache.lucene.search.ScoreMode;
import org.apache.lucene.search.SimpleCollector;
@ -35,9 +37,11 @@ import reactor.util.concurrent.Queues;
public class UnsortedUnscoredContinuousLuceneMultiSearcher implements LuceneMultiSearcher {
private static final Scheduler UNSCORED_UNSORTED_EXECUTOR = Schedulers.newBoundedElastic(Runtime
.getRuntime()
.availableProcessors(), Schedulers.DEFAULT_BOUNDED_ELASTIC_QUEUESIZE, "UnscoredUnsortedExecutor");
private static final Scheduler UNSCORED_UNSORTED_EXECUTOR = Schedulers.newBoundedElastic(
Schedulers.DEFAULT_BOUNDED_ELASTIC_SIZE,
Schedulers.DEFAULT_BOUNDED_ELASTIC_QUEUESIZE,
"UnscoredUnsortedExecutor"
);
private static final Supplier<Queue<ScoreDoc>> QUEUE_SUPPLIER = Queues.get(1024);
@Override
@ -67,35 +71,39 @@ public class UnsortedUnscoredContinuousLuceneMultiSearcher implements LuceneMult
var cm = new CollectorManager<Collector, Void>() {
class IterableCollector extends SimpleCollector {
class IterableCollector implements Collector {
private int shardIndex;
private LeafReaderContext context;
@Override
public void collect(int i) {
if (Schedulers.isInNonBlockingThread()) {
throw new UnsupportedOperationException("Called collect in a nonblocking thread");
}
var scoreDoc = new ScoreDoc(context.docBase + i, 0, shardIndex);
boolean shouldRetry;
do {
var currentError = scoreDocsSink.tryEmitNext(scoreDoc);
shouldRetry = currentError == EmitResult.FAIL_NON_SERIALIZED
|| currentError == EmitResult.FAIL_OVERFLOW
|| currentError == EmitResult.FAIL_ZERO_SUBSCRIBER;
if (shouldRetry) {
LockSupport.parkNanos(10);
}
if (!shouldRetry && currentError.isFailure()) {
currentError.orThrow();
}
} while (shouldRetry);
}
public LeafCollector getLeafCollector(LeafReaderContext leafReaderContext) throws IOException {
return new LeafCollector() {
@Override
public void setScorer(Scorable scorable) throws IOException {
@Override
protected void doSetNextReader(LeafReaderContext context) {
this.context = context;
}
@Override
public void collect(int i) throws IOException {
if (Schedulers.isInNonBlockingThread()) {
throw new UnsupportedOperationException("Called collect in a nonblocking thread");
}
var scoreDoc = new ScoreDoc(leafReaderContext.docBase + i, 0, shardIndex);
boolean shouldRetry;
do {
var currentError = scoreDocsSink.tryEmitNext(scoreDoc);
shouldRetry = currentError == EmitResult.FAIL_NON_SERIALIZED
|| currentError == EmitResult.FAIL_OVERFLOW
|| currentError == EmitResult.FAIL_ZERO_SUBSCRIBER;
if (shouldRetry) {
LockSupport.parkNanos(10);
}
if (!shouldRetry && currentError.isFailure()) {
currentError.orThrow();
}
} while (shouldRetry);
}
};
}
@Override