query
* @return new weight
*/
Index: src/java/org/apache/lucene/search/TopDocCollector.java
===================================================================
--- src/java/org/apache/lucene/search/TopDocCollector.java (revision 575451)
+++ src/java/org/apache/lucene/search/TopDocCollector.java (working copy)
@@ -19,6 +19,8 @@
import org.apache.lucene.util.PriorityQueue;
+import java.io.IOException;
+
/** A {@link HitCollector} implementation that collects the top-scoring
* documents, returning them as a {@link TopDocs}. This is used by {@link
* IndexSearcher} to implement {@link TopDocs}-based search.
@@ -27,7 +29,7 @@
* conditionally invoke super() in order to filter which
* documents are collected.
**/
-public class TopDocCollector extends HitCollector {
+public class TopDocCollector extends TimeLimitedCollector {
private int numHits;
private float minScore = 0.0f;
@@ -41,11 +43,23 @@
this(numHits, new HitQueue(numHits));
}
+ public TopDocCollector(int numHits, int maxTicks, TimerThread timer)
+ throws IOException {
+ this(numHits, new HitQueue(numHits), maxTicks, timer);
+ }
+
TopDocCollector(int numHits, PriorityQueue hq) {
this.numHits = numHits;
this.hq = hq;
}
+ TopDocCollector(int numHits, PriorityQueue hq,
+ int maxTicks, TimerThread timer) {
+ super(maxTicks, timer);
+ this.numHits = numHits;
+ this.hq = hq;
+ }
+
// javadoc inherited
public void collect(int doc, float score) {
if (score > 0.0f) {
@@ -55,8 +69,16 @@
minScore = ((ScoreDoc)hq.top()).score; // maintain minScore
}
}
+ super.collect(doc, score);
}
+ /**
+ * Provide access to super.collect(int, float)
+ */
+ protected void timeLimitedCollect(int doc, float score) {
+ super.collect(doc, score);
+ }
+
/** The total number of documents that matched this query. */
public int getTotalHits() { return totalHits; }
Index: src/java/org/apache/lucene/search/TimerThread.java
===================================================================
--- src/java/org/apache/lucene/search/TimerThread.java (revision 0)
+++ src/java/org/apache/lucene/search/TimerThread.java (revision 0)
@@ -0,0 +1,90 @@
+package org.apache.lucene.search;
+
+/**
+ * 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.
+ */
+
+/**
+ * TimerThread provides a pseudo-clock service to all searching
+ * threads, so that they can count elapsed time with less overhead
+ * than repeatedly calling System.currentTimeMillis. A single
+ * thread should be created to be used for all searches.
+ */
+public class TimerThread extends Thread {
+ private int tick;
+ // NOTE: we can avoid explicit synchronization here for several reasons:
+ // * updates to 32-bit-sized variables are atomic
+ // * only single thread modifies this value
+ // * use of volatile keyword ensures that it does not reside in
+ // a register, but in main memory (so that changes are visible to
+ // other threads).
+ // * visibility of changes does not need to be instantanous, we can
+ // afford losing a tick or two.
+ //
+ // See section 17 of the Java Language Specification for details.
+ public volatile int milliseconds = 0;
+
+ private boolean running = true;
+
+ /**
+ * @param tick The resolution of a single tick in milliseconds.
+ */
+ public TimerThread(int tick) {
+ super("TimeLimitedCollector timer thread");
+ this.tick = tick;
+ this.setDaemon(true);
+ }
+
+ public void run() {
+ boolean interrupted = false;
+ try {
+ while(running) {
+ milliseconds += tick;
+ try {
+ Thread.sleep(tick);
+ } catch (InterruptedException ie) {
+ interrupted = true;
+ // fall through and retry
+ }
+ }
+ } finally {
+ if (interrupted)
+ Thread.currentThread().interrupt();
+ }
+ }
+
+ /**
+ * Get the timer value in milliseconds.
+ */
+ public int getMilliseconds() {
+ return milliseconds;
+ }
+
+ /**
+ * Returns the number of milliseconds elapsed since startTicks. This method
+ * handles an overflow condition.
+ *
+ * @param startMilliseconds The number of millliseconds registered by
+ * getMilliseconds() at the begining of the interval being timed.
+ */
+ public int getElapsedMilliseconds( int startMilliseconds ) {
+ int curMilliseconds = milliseconds;
+ if (curMilliseconds < startMilliseconds) {
+ curMilliseconds += Integer.MAX_VALUE;
+ }
+ return (curMilliseconds - startMilliseconds);
+ }
+}
Index: src/java/org/apache/lucene/search/TopDocs.java
===================================================================
--- src/java/org/apache/lucene/search/TopDocs.java (revision 575451)
+++ src/java/org/apache/lucene/search/TopDocs.java (working copy)
@@ -20,6 +20,7 @@
/** Expert: Returned by low-level search implementations.
* @see Searcher#search(Query,Filter,int) */
public class TopDocs implements java.io.Serializable {
+ private boolean partialResultsFlag = false;
/** Expert: The total number of hits for the query.
* @see Hits#length()
*/
@@ -45,4 +46,22 @@
this.scoreDocs = scoreDocs;
this.maxScore = maxScore;
}
+
+ /**
+ * Indicates that these top hits are based on only looking at some
+ * results. Searching or collecting of the docs was intentionally
+ * limited by some factor such as a time restriction.
+ *
+ * @return true if these top hits are based on only looking at some results.
+ */
+ public boolean arePartialResults() {
+ return partialResultsFlag;
+ }
+
+ /**
+ * Sets partial results flag.
+ */
+ public void setPartialResultsFlag(boolean p) {
+ partialResultsFlag = p;
+ }
}
Index: src/java/org/apache/lucene/search/Hits.java
===================================================================
--- src/java/org/apache/lucene/search/Hits.java (revision 575451)
+++ src/java/org/apache/lucene/search/Hits.java (working copy)
@@ -38,6 +38,7 @@
private Filter filter = null;
private Sort sort = null;
+ private boolean partialResultsFlag = false;
private int length; // the total number of hits
private Vector hitDocs = new Vector(); // cache of hits retrieved
@@ -72,6 +73,7 @@
int n = min * 2; // double # retrieved
TopDocs topDocs = (sort == null) ? searcher.search(weight, filter, n) : searcher.search(weight, filter, n, sort);
+ if( topDocs.arePartialResults() ) { partialResultsFlag = true; }
length = topDocs.totalHits;
ScoreDoc[] scoreDocs = topDocs.scoreDocs;
@@ -145,6 +147,24 @@
return new HitIterator(this);
}
+ /**
+ * Indicates that these top hits are based on only looking at some
+ * results. Searching or collecting of the docs was intentionally
+ * limited by some factor such as a time restriction.
+ *
+ * @return true if these top hits are based on only looking at some results.
+ */
+ public boolean arePartialResults() {
+ return partialResultsFlag;
+ }
+
+ /**
+ * Sets partial results flag.
+ */
+ public void setPartialResultsFlag(boolean p) {
+ partialResultsFlag = p;
+ }
+
private final HitDoc hitDoc(int n) throws IOException {
if (n >= length) {
throw new IndexOutOfBoundsException("Not a valid hit number: " + n);
Index: src/java/org/apache/lucene/search/TimeLimitedCollector.java
===================================================================
--- src/java/org/apache/lucene/search/TimeLimitedCollector.java (revision 0)
+++ src/java/org/apache/lucene/search/TimeLimitedCollector.java (revision 0)
@@ -0,0 +1,74 @@
+package org.apache.lucene.search;
+
+/**
+ * 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.
+ */
+
+/**
+ * The TimeLimitedCollector is used to timeout search requests that + * take longer than the maximum allowed search time limit. After this + * time is exceeded, the search thread is stopped by throwing a + * TimeExceeded Exception.
+ * + *To use the TimeLimitedCollector, your collector should extend + * this class and call super.collect(doc, score);. + * + * @see Searcher#search(Query,HitCollector) + * @version $Id:$ + */ +public class TimeLimitedCollector extends HitCollector { + + private TimerThread timer; + + private int timeAllowed; + private int curMilliseconds; + + public static class TimeExceeded extends RuntimeException { + public long maxTime; + public int maxDoc; + public TimeExceeded(long maxTime, int maxDoc) { + super("Exceeded search time: " + maxTime + " ms."); + this.maxTime = maxTime; + this.maxDoc = maxDoc; + } + } + + /** + * Timer is null and inactive if constructed with default constructor. + */ + public TimeLimitedCollector() { + } + + public TimeLimitedCollector(int timeAllowed, TimerThread timer) { + this.timeAllowed = timeAllowed; + if (timer != null) { + this.timer = timer; + this.curMilliseconds = timer.getMilliseconds(); + } + } + + /** + * @throws TimeExceeded if the time allowed is exceeded. + */ + public void collect(int doc, float score) { + if (timer != null) { + int elapsed = timer.getElapsedMilliseconds( curMilliseconds ); + if (elapsed > timeAllowed) { + throw new TimeExceeded(elapsed, doc); + } + } + } +} Index: src/java/org/apache/lucene/search/IndexSearcher.java =================================================================== --- src/java/org/apache/lucene/search/IndexSearcher.java (revision 575451) +++ src/java/org/apache/lucene/search/IndexSearcher.java (working copy) @@ -109,9 +109,22 @@ if (nDocs <= 0) // null might be returned from hq.top() below. throw new IllegalArgumentException("nDocs must be > 0"); - TopDocCollector collector = new TopDocCollector(nDocs); - search(weight, filter, collector); - return collector.topDocs(); + TopDocCollector collector = new TopDocCollector(nDocs, maxTicks, timer); + TimeLimitedCollector.TimeExceeded te = null; + try { + search(weight, filter, collector); + } + catch( TimeLimitedCollector.TimeExceeded x ) { + te = x; + } + TopDocs topDocs = collector.topDocs(); + if( te != null && topDocs != null ) { + topDocs.setPartialResultsFlag(true); + // Estimate total hits. + topDocs.totalHits = (int)(topDocs.totalHits + * (maxDoc()/(float)te.maxDoc)); + } + return topDocs; } // inherit javadoc @@ -120,12 +133,27 @@ throws IOException { TopFieldDocCollector collector = - new TopFieldDocCollector(reader, sort, nDocs); - search(weight, filter, collector); - return (TopFieldDocs)collector.topDocs(); + new TopFieldDocCollector(reader, sort, nDocs, maxTicks, timer); + TimeLimitedCollector.TimeExceeded te = null; + try { + search(weight, filter, collector); + } + catch( TimeLimitedCollector.TimeExceeded x ) { + te = x; + } + TopFieldDocs topFieldDocs = (TopFieldDocs)collector.topDocs(); + if( te != null && topFieldDocs != null ) { + topFieldDocs.setPartialResultsFlag(true); + // Estimate total hits. + topFieldDocs.totalHits = (int)(topFieldDocs.totalHits + * (maxDoc()/(float)te.maxDoc)); + } + return topFieldDocs; } // inherit javadoc + // FIXME: If HitCollector passed in is a TimeLimitedCollector, + // TimeLimitedCollector.TimeExceeded could be thrown from this method. public void search(Weight weight, Filter filter, final HitCollector results) throws IOException { HitCollector collector = results;