Simplify ContextIndexSearcher.

In particular this commit moves collector wrapping logic from
ContextIndexSearcher to QueryPhase.
This commit is contained in:
Adrien Grand 2015-08-14 13:29:33 +02:00
parent 470f5370b9
commit b3e7146b22
13 changed files with 205 additions and 258 deletions

View File

@ -185,7 +185,7 @@ public class TransportValidateQueryAction extends TransportBroadcastAction<Valid
valid = true;
if (request.explain()) {
explanation = searchContext.query().toString();
explanation = searchContext.parsedQuery().query().toString();
}
if (request.rewrite()) {
explanation = getRewrittenQuery(searcher.searcher(), searchContext.query());

View File

@ -20,9 +20,11 @@ package org.elasticsearch.percolator;
import com.carrotsearch.hppc.ObjectObjectAssociativeContainer;
import com.google.common.collect.ImmutableList;
import org.apache.lucene.index.IndexReader;
import org.apache.lucene.index.IndexableField;
import org.apache.lucene.index.LeafReaderContext;
import org.apache.lucene.search.Collector;
import org.apache.lucene.search.IndexSearcher;
import org.apache.lucene.search.Query;
import org.apache.lucene.search.ScoreDoc;
@ -32,7 +34,6 @@ import org.apache.lucene.util.Counter;
import org.elasticsearch.action.percolate.PercolateShardRequest;
import org.elasticsearch.action.search.SearchType;
import org.elasticsearch.cache.recycler.PageCacheRecycler;
import org.elasticsearch.client.Client;
import org.elasticsearch.common.*;
import org.elasticsearch.common.collect.ImmutableOpenMap;
import org.elasticsearch.common.lease.Releasables;
@ -111,13 +112,13 @@ public class PercolateContext extends SearchContext {
private SearchLookup searchLookup;
private ParsedQuery parsedQuery;
private Query query;
private boolean queryRewritten;
private Query percolateQuery;
private FetchSubPhase.HitContext hitContext;
private SearchContextAggregations aggregations;
private QuerySearchResult querySearchResult;
private Sort sort;
private final Map<String, FetchSubPhaseContext> subPhaseContexts = new HashMap<>();
private final Map<Class<?>, Collector> queryCollectors = new HashMap<>();
public PercolateContext(PercolateShardRequest request, SearchShardTarget searchShardTarget, IndexShard indexShard,
IndexService indexService, PageCacheRecycler pageCacheRecycler,
@ -232,7 +233,6 @@ public class PercolateContext extends SearchContext {
public SearchContext parsedQuery(ParsedQuery query) {
this.parsedQuery = query;
this.query = query.query();
this.queryRewritten = false;
return this;
}
@ -246,18 +246,6 @@ public class PercolateContext extends SearchContext {
return query;
}
@Override
public boolean queryRewritten() {
return queryRewritten;
}
@Override
public SearchContext updateRewriteQuery(Query rewriteQuery) {
queryRewritten = true;
query = rewriteQuery;
return this;
}
@Override
public String[] types() {
return types;
@ -768,4 +756,9 @@ public class PercolateContext extends SearchContext {
public void copyContextAndHeadersFrom(HasContextAndHeaders other) {
assert false : "percolatecontext does not support contexts & headers";
}
@Override
public Map<Class<?>, Collector> queryCollectors() {
return queryCollectors;
}
}

View File

@ -86,7 +86,7 @@ public class AggregationPhase implements SearchPhase {
if (!collectors.isEmpty()) {
final BucketCollector collector = BucketCollector.wrap(collectors);
collector.preCollection();
context.searcher().queryCollectors().put(AggregationPhase.class, collector);
context.queryCollectors().put(AggregationPhase.class, collector);
}
} catch (IOException e) {
throw new AggregationInitializationException("Could not initialize aggregators", e);
@ -162,7 +162,7 @@ public class AggregationPhase implements SearchPhase {
// disable aggregations so that they don't run on next pages in case of scrolling
context.aggregations(null);
context.searcher().queryCollectors().remove(AggregationPhase.class);
context.queryCollectors().remove(AggregationPhase.class);
}
}

View File

@ -57,10 +57,6 @@ public class DfsPhase implements SearchPhase {
public void execute(SearchContext context) {
final ObjectHashSet<Term> termsSet = new ObjectHashSet<>();
try {
if (!context.queryRewritten()) {
context.updateRewriteQuery(context.searcher().rewrite(context.query()));
}
context.searcher().createNormalizedWeight(context.query(), true).extractTerms(new DelegateSet(termsSet));
for (RescoreSearchContext rescoreContext : context.rescore()) {
rescoreContext.rescorer().extractTerms(context, rescoreContext, new DelegateSet(termsSet));

View File

@ -20,41 +20,25 @@
package org.elasticsearch.search.internal;
import org.apache.lucene.index.LeafReaderContext;
import org.apache.lucene.search.BooleanClause.Occur;
import org.apache.lucene.search.BooleanQuery;
import org.apache.lucene.search.Collector;
import org.apache.lucene.search.Explanation;
import org.apache.lucene.search.IndexSearcher;
import org.apache.lucene.search.MultiCollector;
import org.apache.lucene.search.Query;
import org.apache.lucene.search.TimeLimitingCollector;
import org.apache.lucene.search.Weight;
import org.elasticsearch.ExceptionsHelper;
import org.elasticsearch.common.lease.Releasable;
import org.elasticsearch.common.lucene.Lucene;
import org.elasticsearch.common.lucene.MinimumScoreCollector;
import org.elasticsearch.common.lucene.search.FilteredCollector;
import org.elasticsearch.index.engine.Engine;
import org.elasticsearch.search.SearchService;
import org.elasticsearch.search.dfs.CachedDfSource;
import org.elasticsearch.search.internal.SearchContext.Lifetime;
import java.io.IOException;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
/**
* Context-aware extension of {@link IndexSearcher}.
*/
public class ContextIndexSearcher extends IndexSearcher implements Releasable {
public static enum Stage {
NA,
MAIN_QUERY
}
/** The wrapped {@link IndexSearcher}. The reason why we sometimes prefer delegating to this searcher instead of <tt>super</tt> is that
* this instance may have more assertions, for example if it comes from MockInternalEngine which wraps the IndexSearcher into an
* AssertingIndexSearcher. */
@ -64,10 +48,6 @@ public class ContextIndexSearcher extends IndexSearcher implements Releasable {
private CachedDfSource dfSource;
private Map<Class<?>, Collector> queryCollectors;
private Stage currentState = Stage.NA;
public ContextIndexSearcher(SearchContext searchContext, Engine.Searcher searcher) {
super(searcher.reader());
in = searcher.searcher();
@ -83,49 +63,11 @@ public class ContextIndexSearcher extends IndexSearcher implements Releasable {
this.dfSource = dfSource;
}
/**
* Adds a query level collector that runs at {@link Stage#MAIN_QUERY}. Note, supports
* {@link org.elasticsearch.common.lucene.search.XCollector} allowing for a callback
* when collection is done.
*/
public Map<Class<?>, Collector> queryCollectors() {
if (queryCollectors == null) {
queryCollectors = new HashMap<>();
}
return queryCollectors;
}
public void inStage(Stage stage) {
this.currentState = stage;
}
public void finishStage(Stage stage) {
assert currentState == stage : "Expected stage " + stage + " but was stage " + currentState;
this.currentState = Stage.NA;
}
@Override
public Query rewrite(Query original) throws IOException {
if (original == searchContext.query() || original == searchContext.parsedQuery().query()) {
// optimize in case its the top level search query and we already rewrote it...
if (searchContext.queryRewritten()) {
return searchContext.query();
}
Query rewriteQuery = in.rewrite(original);
searchContext.updateRewriteQuery(rewriteQuery);
return rewriteQuery;
} else {
return in.rewrite(original);
}
}
@Override
public Weight createNormalizedWeight(Query query, boolean needsScores) throws IOException {
// TODO: needsScores
// can we avoid dfs stuff here if we dont need scores?
try {
// if its the main query, use we have dfs data, only then do it
if (dfSource != null && (query == searchContext.query() || query == searchContext.parsedQuery().query())) {
// if scores are needed and we have dfs data then use it
if (dfSource != null && needsScores) {
return dfSource.createNormalizedWeight(query, needsScores);
}
return in.createNormalizedWeight(query, needsScores);
@ -135,81 +77,19 @@ public class ContextIndexSearcher extends IndexSearcher implements Releasable {
}
}
@Override
public void search(Query query, Collector collector) throws IOException {
// Wrap the caller's collector with various wrappers e.g. those used to siphon
// matches off for aggregation or to impose a time-limit on collection.
final boolean timeoutSet = searchContext.timeoutInMillis() != SearchService.NO_TIMEOUT.millis();
final boolean terminateAfterSet = searchContext.terminateAfter() != SearchContext.DEFAULT_TERMINATE_AFTER;
if (timeoutSet) {
// TODO: change to use our own counter that uses the scheduler in ThreadPool
// throws TimeLimitingCollector.TimeExceededException when timeout has reached
collector = Lucene.wrapTimeLimitingCollector(collector, searchContext.timeEstimateCounter(), searchContext.timeoutInMillis());
}
if (terminateAfterSet) {
// throws Lucene.EarlyTerminationException when given count is reached
collector = Lucene.wrapCountBasedEarlyTerminatingCollector(collector, searchContext.terminateAfter());
}
if (currentState == Stage.MAIN_QUERY) {
if (searchContext.parsedPostFilter() != null) {
// this will only get applied to the actual search collector and not
// to any scoped collectors, also, it will only be applied to the main collector
// since that is where the filter should only work
final Weight filterWeight = createNormalizedWeight(searchContext.parsedPostFilter().query(), false);
collector = new FilteredCollector(collector, filterWeight);
}
if (queryCollectors != null && !queryCollectors.isEmpty()) {
ArrayList<Collector> allCollectors = new ArrayList<>(queryCollectors.values());
allCollectors.add(collector);
collector = MultiCollector.wrap(allCollectors);
}
// apply the minimum score after multi collector so we filter aggs as well
if (searchContext.minimumScore() != null) {
collector = new MinimumScoreCollector(collector, searchContext.minimumScore());
}
}
super.search(query, collector);
}
@Override
public void search(List<LeafReaderContext> leaves, Weight weight, Collector collector) throws IOException {
final boolean timeoutSet = searchContext.timeoutInMillis() != SearchService.NO_TIMEOUT.millis();
final boolean terminateAfterSet = searchContext.terminateAfter() != SearchContext.DEFAULT_TERMINATE_AFTER;
public Explanation explain(Query query, int doc) throws IOException {
try {
if (timeoutSet || terminateAfterSet) {
try {
super.search(leaves, weight, collector);
} catch (TimeLimitingCollector.TimeExceededException e) {
assert timeoutSet : "TimeExceededException thrown even though timeout wasn't set";
searchContext.queryResult().searchTimedOut(true);
} catch (Lucene.EarlyTerminationException e) {
assert terminateAfterSet : "EarlyTerminationException thrown even though terminateAfter wasn't set";
searchContext.queryResult().terminatedEarly(true);
}
if (terminateAfterSet && searchContext.queryResult().terminatedEarly() == null) {
searchContext.queryResult().terminatedEarly(false);
}
} else {
super.search(leaves, weight, collector);
}
return in.explain(query, doc);
} finally {
searchContext.clearReleasables(Lifetime.COLLECTION);
}
}
@Override
public Explanation explain(Query query, int doc) throws IOException {
protected void search(List<LeafReaderContext> leaves, Weight weight, Collector collector) throws IOException {
try {
if (searchContext.aliasFilter() == null) {
return super.explain(query, doc);
}
BooleanQuery filteredQuery = new BooleanQuery();
filteredQuery.add(query, Occur.MUST);
filteredQuery.add(searchContext.aliasFilter(), Occur.FILTER);
return super.explain(filteredQuery, doc);
super.search(leaves, weight, collector);
} finally {
searchContext.clearReleasables(Lifetime.COLLECTION);
}

View File

@ -22,6 +22,7 @@ package org.elasticsearch.search.internal;
import com.carrotsearch.hppc.ObjectObjectAssociativeContainer;
import com.google.common.collect.ImmutableList;
import com.google.common.collect.Lists;
import org.apache.lucene.search.BooleanClause.Occur;
import org.apache.lucene.search.*;
import org.apache.lucene.util.Counter;
@ -66,6 +67,7 @@ import org.elasticsearch.search.rescore.RescoreSearchContext;
import org.elasticsearch.search.scan.ScanContext;
import org.elasticsearch.search.suggest.SuggestionSearchContext;
import java.io.IOException;
import java.util.*;
/**
@ -119,7 +121,6 @@ public class DefaultSearchContext extends SearchContext {
private SuggestionSearchContext suggest;
private List<RescoreSearchContext> rescore;
private SearchLookup searchLookup;
private boolean queryRewritten;
private volatile long keepAlive;
private ScoreDoc lastEmittedDoc;
private final long originNanoTime = System.nanoTime();
@ -127,6 +128,7 @@ public class DefaultSearchContext extends SearchContext {
private InnerHitsContext innerHitsContext;
private final Map<String, FetchSubPhaseContext> subPhaseContexts = new HashMap<>();
private final Map<Class<?>, Collector> queryCollectors = new HashMap<>();
public DefaultSearchContext(long id, ShardSearchRequest request, SearchShardTarget shardTarget,
Engine.Searcher engineSearcher, IndexService indexService, IndexShard indexShard,
@ -197,10 +199,15 @@ public class DefaultSearchContext extends SearchContext {
parsedQuery(new ParsedQuery(filtered, parsedQuery()));
}
}
try {
this.query = searcher().rewrite(this.query);
} catch (IOException e) {
throw new QueryPhaseExecutionException(this, "Failed to rewrite main query", e);
}
}
@Override
public Filter searchFilter(String[] types) {
public Query searchFilter(String[] types) {
Query filter = mapperService().searchFilter(types);
if (filter == null && aliasFilter == null) {
return null;
@ -212,7 +219,7 @@ public class DefaultSearchContext extends SearchContext {
if (aliasFilter != null) {
bq.add(aliasFilter, Occur.MUST);
}
return new QueryWrapperFilter(bq);
return new ConstantScoreQuery(bq);
}
@Override
@ -513,7 +520,6 @@ public class DefaultSearchContext extends SearchContext {
@Override
public SearchContext parsedQuery(ParsedQuery query) {
queryRewritten = false;
this.originalQuery = query;
this.query = query.query();
return this;
@ -525,31 +531,13 @@ public class DefaultSearchContext extends SearchContext {
}
/**
* The query to execute, might be rewritten.
* The query to execute, in its rewritten form.
*/
@Override
public Query query() {
return this.query;
}
/**
* Has the query been rewritten already?
*/
@Override
public boolean queryRewritten() {
return queryRewritten;
}
/**
* Rewrites the query and updates it. Only happens once.
*/
@Override
public SearchContext updateRewriteQuery(Query rewriteQuery) {
query = rewriteQuery;
queryRewritten = true;
return this;
}
@Override
public int from() {
return from;
@ -810,4 +798,9 @@ public class DefaultSearchContext extends SearchContext {
public void copyContextAndHeadersFrom(HasContextAndHeaders other) {
request.copyContextAndHeadersFrom(other);
}
@Override
public Map<Class<?>, Collector> queryCollectors() {
return queryCollectors;
}
}

View File

@ -20,6 +20,8 @@
package org.elasticsearch.search.internal;
import com.carrotsearch.hppc.ObjectObjectAssociativeContainer;
import org.apache.lucene.search.Collector;
import org.apache.lucene.search.Query;
import org.apache.lucene.search.ScoreDoc;
import org.apache.lucene.search.Sort;
@ -58,6 +60,7 @@ import org.elasticsearch.search.scan.ScanContext;
import org.elasticsearch.search.suggest.SuggestionSearchContext;
import java.util.List;
import java.util.Map;
import java.util.Set;
public abstract class FilteredSearchContext extends SearchContext {
@ -375,16 +378,6 @@ public abstract class FilteredSearchContext extends SearchContext {
return in.query();
}
@Override
public boolean queryRewritten() {
return in.queryRewritten();
}
@Override
public SearchContext updateRewriteQuery(Query rewriteQuery) {
return in.updateRewriteQuery(rewriteQuery);
}
@Override
public int from() {
return in.from();
@ -624,4 +617,9 @@ public abstract class FilteredSearchContext extends SearchContext {
public <SubPhaseContext extends FetchSubPhaseContext> SubPhaseContext getFetchSubPhaseContext(FetchSubPhase.ContextFactory<SubPhaseContext> contextFactory) {
return in.getFetchSubPhaseContext(contextFactory);
}
@Override
public Map<Class<?>, Collector> queryCollectors() {
return in.queryCollectors();
}
}

View File

@ -21,6 +21,8 @@ package org.elasticsearch.search.internal;
import com.google.common.collect.Iterables;
import com.google.common.collect.Multimap;
import com.google.common.collect.MultimapBuilder;
import org.apache.lucene.search.Collector;
import org.apache.lucene.search.Query;
import org.apache.lucene.search.ScoreDoc;
import org.apache.lucene.search.Sort;
@ -65,6 +67,7 @@ import org.elasticsearch.search.suggest.SuggestionSearchContext;
import java.util.ArrayList;
import java.util.Collection;
import java.util.List;
import java.util.Map;
import java.util.concurrent.atomic.AtomicBoolean;
public abstract class SearchContext implements Releasable, HasContextAndHeaders {
@ -257,16 +260,6 @@ public abstract class SearchContext implements Releasable, HasContextAndHeaders
*/
public abstract Query query();
/**
* Has the query been rewritten already?
*/
public abstract boolean queryRewritten();
/**
* Rewrites the query and updates it. Only happens once.
*/
public abstract SearchContext updateRewriteQuery(Query rewriteQuery);
public abstract int from();
public abstract SearchContext from(int from);
@ -359,6 +352,9 @@ public abstract class SearchContext implements Releasable, HasContextAndHeaders
public abstract Counter timeEstimateCounter();
/** Return a view of the additional query collectors that should be run for this context. */
public abstract Map<Class<?>, Collector> queryCollectors();
/**
* The life time of an object that is used during search execution.
*/

View File

@ -206,11 +206,6 @@ public class SubSearchContext extends FilteredSearchContext {
throw new UnsupportedOperationException("Not supported");
}
@Override
public SearchContext updateRewriteQuery(Query rewriteQuery) {
throw new UnsupportedOperationException("Not supported");
}
@Override
public int from() {
return from;

View File

@ -21,24 +21,42 @@ package org.elasticsearch.search.query;
import com.google.common.collect.ImmutableMap;
import org.apache.lucene.search.Collector;
import org.apache.lucene.search.FieldDoc;
import org.apache.lucene.search.IndexSearcher;
import org.apache.lucene.search.MultiCollector;
import org.apache.lucene.search.Query;
import org.apache.lucene.search.ScoreDoc;
import org.apache.lucene.search.TimeLimitingCollector;
import org.apache.lucene.search.TopDocs;
import org.apache.lucene.search.TopDocsCollector;
import org.apache.lucene.search.TopFieldCollector;
import org.apache.lucene.search.TopScoreDocCollector;
import org.apache.lucene.search.TotalHitCountCollector;
import org.apache.lucene.search.Weight;
import org.elasticsearch.action.search.SearchType;
import org.elasticsearch.common.inject.Inject;
import org.elasticsearch.common.lucene.Lucene;
import org.elasticsearch.common.lucene.MinimumScoreCollector;
import org.elasticsearch.common.lucene.search.FilteredCollector;
import org.elasticsearch.search.SearchParseElement;
import org.elasticsearch.search.SearchPhase;
import org.elasticsearch.search.SearchService;
import org.elasticsearch.search.aggregations.AggregationPhase;
import org.elasticsearch.search.internal.ContextIndexSearcher;
import org.elasticsearch.search.internal.SearchContext;
import org.elasticsearch.search.internal.SearchContext.Lifetime;
import org.elasticsearch.search.rescore.RescorePhase;
import org.elasticsearch.search.rescore.RescoreSearchContext;
import org.elasticsearch.search.scan.ScanContext.ScanCollector;
import org.elasticsearch.search.sort.SortParseElement;
import org.elasticsearch.search.sort.TrackScoresParseElement;
import org.elasticsearch.search.suggest.SuggestPhase;
import java.io.IOException;
import java.util.ArrayList;
import java.util.List;
import java.util.Map;
import java.util.concurrent.Callable;
/**
*
@ -97,39 +115,120 @@ public class QueryPhase implements SearchPhase {
searchContext.queryResult().searchTimedOut(false);
searchContext.searcher().inStage(ContextIndexSearcher.Stage.MAIN_QUERY);
boolean rescore = false;
try {
searchContext.queryResult().from(searchContext.from());
searchContext.queryResult().size(searchContext.size());
final IndexSearcher searcher = searchContext.searcher();
Query query = searchContext.query();
final TopDocs topDocs;
int numDocs = searchContext.from() + searchContext.size();
final int totalNumDocs = searcher.getIndexReader().numDocs();
int numDocs = Math.min(searchContext.from() + searchContext.size(), totalNumDocs);
Collector collector;
final Callable<TopDocs> topDocsCallable;
if (searchContext.size() == 0) { // no matter what the value of from is
topDocs = new TopDocs(searchContext.searcher().count(query), Lucene.EMPTY_SCORE_DOCS, 0);
final TotalHitCountCollector totalHitCountCollector = new TotalHitCountCollector();
collector = totalHitCountCollector;
topDocsCallable = new Callable<TopDocs>() {
@Override
public TopDocs call() throws Exception {
return new TopDocs(totalHitCountCollector.getTotalHits(), Lucene.EMPTY_SCORE_DOCS, 0);
}
};
} else if (searchContext.searchType() == SearchType.SCAN) {
topDocs = searchContext.scanContext().execute(searchContext);
query = searchContext.scanContext().wrapQuery(query);
final ScanCollector scanCollector = searchContext.scanContext().collector(searchContext);
collector = scanCollector;
topDocsCallable = new Callable<TopDocs>() {
@Override
public TopDocs call() throws Exception {
return scanCollector.topDocs();
}
};
} else {
// Perhaps have a dedicated scroll phase?
final TopDocsCollector<?> topDocsCollector;
ScoreDoc lastEmittedDoc;
if (searchContext.request().scroll() != null) {
numDocs = searchContext.size();
ScoreDoc lastEmittedDoc = searchContext.lastEmittedDoc();
numDocs = Math.min(searchContext.size(), totalNumDocs);
lastEmittedDoc = searchContext.lastEmittedDoc();
} else {
lastEmittedDoc = null;
}
if (totalNumDocs == 0) {
// top collectors don't like a size of 0
numDocs = 1;
}
assert numDocs > 0;
if (searchContext.sort() != null) {
topDocs = searchContext.searcher().searchAfter(
lastEmittedDoc, query, null, numDocs, searchContext.sort(),
searchContext.trackScores(), searchContext.trackScores()
);
topDocsCollector = TopFieldCollector.create(searchContext.sort(), numDocs,
(FieldDoc) lastEmittedDoc, true, searchContext.trackScores(), searchContext.trackScores());
} else {
rescore = !searchContext.rescore().isEmpty();
for (RescoreSearchContext rescoreContext : searchContext.rescore()) {
numDocs = Math.max(rescoreContext.window(), numDocs);
}
topDocs = searchContext.searcher().searchAfter(lastEmittedDoc, query, numDocs);
topDocsCollector = TopScoreDocCollector.create(numDocs, lastEmittedDoc);
}
collector = topDocsCollector;
topDocsCallable = new Callable<TopDocs>() {
@Override
public TopDocs call() throws Exception {
return topDocsCollector.topDocs();
}
};
}
final boolean terminateAfterSet = searchContext.terminateAfter() != SearchContext.DEFAULT_TERMINATE_AFTER;
if (terminateAfterSet) {
// throws Lucene.EarlyTerminationException when given count is reached
collector = Lucene.wrapCountBasedEarlyTerminatingCollector(collector, searchContext.terminateAfter());
}
if (searchContext.parsedPostFilter() != null) {
// this will only get applied to the actual search collector and not
// to any scoped collectors, also, it will only be applied to the main collector
// since that is where the filter should only work
final Weight filterWeight = searcher.createNormalizedWeight(searchContext.parsedPostFilter().query(), false);
collector = new FilteredCollector(collector, filterWeight);
}
// plug in additional collectors, like aggregations
List<Collector> allCollectors = new ArrayList<>();
allCollectors.add(collector);
allCollectors.addAll(searchContext.queryCollectors().values());
collector = MultiCollector.wrap(allCollectors);
// apply the minimum score after multi collector so we filter aggs as well
if (searchContext.minimumScore() != null) {
collector = new MinimumScoreCollector(collector, searchContext.minimumScore());
}
final boolean timeoutSet = searchContext.timeoutInMillis() != SearchService.NO_TIMEOUT.millis();
if (timeoutSet) {
// TODO: change to use our own counter that uses the scheduler in ThreadPool
// throws TimeLimitingCollector.TimeExceededException when timeout has reached
collector = Lucene.wrapTimeLimitingCollector(collector, searchContext.timeEstimateCounter(), searchContext.timeoutInMillis());
}
try {
searcher.search(query, collector);
} catch (TimeLimitingCollector.TimeExceededException e) {
assert timeoutSet : "TimeExceededException thrown even though timeout wasn't set";
searchContext.queryResult().searchTimedOut(true);
} catch (Lucene.EarlyTerminationException e) {
assert terminateAfterSet : "EarlyTerminationException thrown even though terminateAfter wasn't set";
searchContext.queryResult().terminatedEarly(true);
}
if (terminateAfterSet && searchContext.queryResult().terminatedEarly() == null) {
searchContext.queryResult().terminatedEarly(false);
}
final TopDocs topDocs = topDocsCallable.call();
if (searchContext.request().scroll() != null) {
int size = topDocs.scoreDocs.length;
if (size > 0) {
// In the case of *QUERY_AND_FETCH we don't get back to shards telling them which least
@ -139,24 +238,10 @@ public class QueryPhase implements SearchPhase {
searchContext.lastEmittedDoc(topDocs.scoreDocs[size - 1]);
}
}
} else {
if (searchContext.sort() != null) {
topDocs = searchContext.searcher().search(query, null, numDocs, searchContext.sort(),
searchContext.trackScores(), searchContext.trackScores());
} else {
rescore = !searchContext.rescore().isEmpty();
for (RescoreSearchContext rescoreContext : searchContext.rescore()) {
numDocs = Math.max(rescoreContext.window(), numDocs);
}
topDocs = searchContext.searcher().search(query, numDocs);
}
}
}
searchContext.queryResult().topDocs(topDocs);
} catch (Throwable e) {
throw new QueryPhaseExecutionException(searchContext, "Failed to execute main query", e);
} finally {
searchContext.searcher().finishStage(ContextIndexSearcher.Stage.MAIN_QUERY);
}
if (rescore) { // only if we do a regular search
rescorePhase.execute(searchContext);

View File

@ -47,18 +47,23 @@ public class ScanContext {
private volatile int docUpTo;
public TopDocs execute(SearchContext context) throws IOException {
return execute(context.searcher(), context.query(), context.size(), context.trackScores());
public ScanCollector collector(SearchContext context) {
return collector(context.size(), context.trackScores());
}
TopDocs execute(IndexSearcher searcher, Query query, int size, boolean trackScores) throws IOException {
ScanCollector collector = new ScanCollector(size, trackScores);
Query q = Queries.filtered(query, new MinDocQuery(docUpTo));
searcher.search(q, collector);
return collector.topDocs();
/** Create a {@link ScanCollector} for the given page size. */
ScanCollector collector(int size, boolean trackScores) {
return new ScanCollector(size, trackScores);
}
private class ScanCollector extends SimpleCollector {
/**
* Wrap the query so that it can skip directly to the right document.
*/
public Query wrapQuery(Query query) {
return Queries.filtered(query, new MinDocQuery(docUpTo));
}
public class ScanCollector extends SimpleCollector {
private final List<ScoreDoc> docs;
@ -70,7 +75,7 @@ public class ScanContext {
private int docBase;
ScanCollector(int size, boolean trackScores) {
private ScanCollector(int size, boolean trackScores) {
this.trackScores = trackScores;
this.docs = new ArrayList<>(size);
this.size = size;

View File

@ -31,8 +31,10 @@ import org.apache.lucene.search.QueryUtils;
import org.apache.lucene.search.ScoreDoc;
import org.apache.lucene.search.Sort;
import org.apache.lucene.search.TermQuery;
import org.apache.lucene.search.TopDocs;
import org.apache.lucene.store.Directory;
import org.elasticsearch.search.scan.ScanContext.MinDocQuery;
import org.elasticsearch.search.scan.ScanContext.ScanCollector;
import org.elasticsearch.test.ESTestCase;
import java.io.IOException;
@ -69,6 +71,13 @@ public class ScanContextTests extends ESTestCase {
dir.close();
}
private static TopDocs execute(IndexSearcher searcher, ScanContext ctx, Query query, int pageSize, boolean trackScores) throws IOException {
query = ctx.wrapQuery(query);
ScanCollector collector = ctx.collector(pageSize, trackScores);
searcher.search(query, collector);
return collector.topDocs();
}
public void testRandom() throws Exception {
final int numDocs = randomIntBetween(10, 200);
final Document doc1 = new Document();
@ -93,10 +102,10 @@ public class ScanContextTests extends ESTestCase {
final List<ScoreDoc> actual = new ArrayList<>();
ScanContext context = new ScanContext();
while (true) {
final ScoreDoc[] page = context.execute(searcher, query, pageSize, trackScores).scoreDocs;
final ScoreDoc[] page = execute(searcher,context, query, pageSize, trackScores).scoreDocs;
assertTrue(page.length <= pageSize);
if (page.length == 0) {
assertEquals(0, context.execute(searcher, query, pageSize, trackScores).scoreDocs.length);
assertEquals(0, execute(searcher, context, query, pageSize, trackScores).scoreDocs.length);
break;
}
actual.addAll(Arrays.asList(page));

View File

@ -20,6 +20,7 @@ package org.elasticsearch.test;
import com.carrotsearch.hppc.ObjectObjectAssociativeContainer;
import org.apache.lucene.search.Collector;
import org.apache.lucene.search.Filter;
import org.apache.lucene.search.Query;
import org.apache.lucene.search.ScoreDoc;
@ -74,6 +75,7 @@ public class TestSearchContext extends SearchContext {
final IndexFieldDataService indexFieldDataService;
final BitsetFilterCache fixedBitSetFilterCache;
final ThreadPool threadPool;
final Map<Class<?>, Collector> queryCollectors = new HashMap<>();
ContextIndexSearcher searcher;
int size;
@ -410,16 +412,6 @@ public class TestSearchContext extends SearchContext {
return null;
}
@Override
public boolean queryRewritten() {
return false;
}
@Override
public SearchContext updateRewriteQuery(Query rewriteQuery) {
return null;
}
@Override
public int from() {
return 0;
@ -667,4 +659,9 @@ public class TestSearchContext extends SearchContext {
@Override
public void copyContextAndHeadersFrom(HasContextAndHeaders other) {}
@Override
public Map<Class<?>, Collector> queryCollectors() {
return queryCollectors;
}
}