You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@lucene.apache.org by sh...@apache.org on 2013/04/09 16:28:53 UTC
svn commit: r1466057 - in /lucene/dev/branches/branch_4x: ./ lucene/
lucene/core/ lucene/core/src/java/org/apache/lucene/index/
lucene/core/src/java/org/apache/lucene/search/
lucene/core/src/test/org/apache/lucene/search/ lucene/misc/
lucene/misc/src/j...
Author: shaie
Date: Tue Apr 9 14:28:52 2013
New Revision: 1466057
URL: http://svn.apache.org/r1466057
Log:
LUCENE-4858: add EarlyTerminatingSortingCollector
Added:
lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/EarlyTerminatingSortingCollector.java
- copied unchanged from r1466047, lucene/dev/trunk/lucene/misc/src/java/org/apache/lucene/index/sorter/EarlyTerminatingSortingCollector.java
lucene/dev/branches/branch_4x/lucene/misc/src/test/org/apache/lucene/index/sorter/TestEarlyTermination.java
- copied unchanged from r1466047, lucene/dev/trunk/lucene/misc/src/test/org/apache/lucene/index/sorter/TestEarlyTermination.java
Modified:
lucene/dev/branches/branch_4x/ (props changed)
lucene/dev/branches/branch_4x/lucene/ (props changed)
lucene/dev/branches/branch_4x/lucene/CHANGES.txt (contents, props changed)
lucene/dev/branches/branch_4x/lucene/core/ (props changed)
lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/index/DocumentsWriterPerThread.java
lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/index/IndexWriter.java
lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/index/MergePolicy.java
lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/index/SegmentReader.java
lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/search/IndexSearcher.java
lucene/dev/branches/branch_4x/lucene/core/src/test/org/apache/lucene/search/TestEarlyTermination.java
lucene/dev/branches/branch_4x/lucene/misc/ (props changed)
lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/NumericDocValuesSorter.java
lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/Sorter.java
lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/SortingAtomicReader.java
lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/SortingMergePolicy.java
lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/package.html
lucene/dev/branches/branch_4x/lucene/misc/src/test/org/apache/lucene/index/sorter/SortingAtomicReaderTest.java
lucene/dev/branches/branch_4x/lucene/misc/src/test/org/apache/lucene/index/sorter/TestSortingMergePolicy.java
Modified: lucene/dev/branches/branch_4x/lucene/CHANGES.txt
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_4x/lucene/CHANGES.txt?rev=1466057&r1=1466056&r2=1466057&view=diff
==============================================================================
--- lucene/dev/branches/branch_4x/lucene/CHANGES.txt (original)
+++ lucene/dev/branches/branch_4x/lucene/CHANGES.txt Tue Apr 9 14:28:52 2013
@@ -146,6 +146,10 @@ New Features
* LUCENE-4902: Added FilterDirectoryReader to allow easy filtering of a
DirectoryReader's subreaders. (Alan Woodward, Adrien Grand, Uwe Schindler)
+* LUCENE-4858: Added EarlyTerminatingSortingCollector to be used in conjunction
+ with SortingMergePolicy, which allows to early terminate queries on sorted
+ indexes, when the sort order matches the index order. (Adrien Grand, Shai Erera)
+
Optimizations
* LUCENE-4839: SorterTemplate.merge can now be overridden in order to replace
Modified: lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/index/DocumentsWriterPerThread.java
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/index/DocumentsWriterPerThread.java?rev=1466057&r1=1466056&r2=1466057&view=diff
==============================================================================
--- lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/index/DocumentsWriterPerThread.java (original)
+++ lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/index/DocumentsWriterPerThread.java Tue Apr 9 14:28:52 2013
@@ -554,7 +554,7 @@ class DocumentsWriterPerThread {
SegmentInfoPerCommit newSegment = flushedSegment.segmentInfo;
- IndexWriter.setDiagnostics(newSegment.info, "flush");
+ IndexWriter.setDiagnostics(newSegment.info, IndexWriter.SOURCE_FLUSH);
IOContext context = new IOContext(new FlushInfo(newSegment.info.getDocCount(), newSegment.sizeInBytes()));
Modified: lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/index/IndexWriter.java
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/index/IndexWriter.java?rev=1466057&r1=1466056&r2=1466057&view=diff
==============================================================================
--- lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/index/IndexWriter.java (original)
+++ lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/index/IndexWriter.java Tue Apr 9 14:28:52 2013
@@ -187,13 +187,21 @@ import org.apache.lucene.util.ThreadInte
public class IndexWriter implements Closeable, TwoPhaseCommit {
private static final int UNBOUNDED_MAX_MERGE_SEGMENTS = -1;
-
/**
* Name of the write lock in the index.
*/
public static final String WRITE_LOCK_NAME = "write.lock";
+ /** Key for the source of a segment in the {@link SegmentInfo#getDiagnostics() diagnostics}. */
+ public static final String SOURCE = "source";
+ /** Source of a segment which results from a merge of other segments. */
+ public static final String SOURCE_MERGE = "merge";
+ /** Source of a segment which results from a flush. */
+ public static final String SOURCE_FLUSH = "flush";
+ /** Source of a segment which results from a call to {@link #addIndexes(IndexReader...)}. */
+ public static final String SOURCE_ADDINDEXES_READERS = "addIndexes(IndexReader...)";
+
/**
* Absolute hard maximum length for a term, in bytes once
* encoded as UTF8. If a term arrives from the analyzer
@@ -2435,7 +2443,7 @@ public class IndexWriter implements Clos
info.setFiles(new HashSet<String>(trackingDir.getCreatedFiles()));
trackingDir.getCreatedFiles().clear();
- setDiagnostics(info, "addIndexes(IndexReader...)");
+ setDiagnostics(info, SOURCE_ADDINDEXES_READERS);
boolean useCompoundFile;
synchronized(this) { // Guard segmentInfos
@@ -3500,16 +3508,15 @@ public class IndexWriter implements Clos
// names.
final String mergeSegmentName = newSegmentName();
SegmentInfo si = new SegmentInfo(directory, Constants.LUCENE_MAIN_VERSION, mergeSegmentName, -1, false, codec, null, null);
- merge.info = new SegmentInfoPerCommit(si, 0, -1L);
+ Map<String,String> details = new HashMap<String,String>();
+ details.put("mergeMaxNumSegments", "" + merge.maxNumSegments);
+ details.put("mergeFactor", Integer.toString(merge.segments.size()));
+ setDiagnostics(si, SOURCE_MERGE, details);
+ merge.setInfo(new SegmentInfoPerCommit(si, 0, -1L));
// Lock order: IW -> BD
bufferedDeletesStream.prune(segmentInfos);
- Map<String,String> details = new HashMap<String,String>();
- details.put("mergeMaxNumSegments", ""+merge.maxNumSegments);
- details.put("mergeFactor", Integer.toString(merge.segments.size()));
- setDiagnostics(si, "merge", details);
-
if (infoStream.isEnabled("IW")) {
infoStream.message("IW", "merge seg=" + merge.info.info.name + " " + segString(merge.segments));
}
Modified: lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/index/MergePolicy.java
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/index/MergePolicy.java?rev=1466057&r1=1466056&r2=1466057&view=diff
==============================================================================
--- lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/index/MergePolicy.java (original)
+++ lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/index/MergePolicy.java Tue Apr 9 14:28:52 2013
@@ -149,6 +149,14 @@ public abstract class MergePolicy implem
}
return Collections.unmodifiableList(readers);
}
+
+ /**
+ * Expert: Sets the {@link SegmentInfoPerCommit} of this {@link OneMerge}.
+ * Allows sub-classes to e.g. set diagnostics properties.
+ */
+ public void setInfo(SegmentInfoPerCommit info) {
+ this.info = info;
+ }
/** Expert: If {@link #getMergeReaders()} reorders document IDs, this method
* must be overridden to return a mapping from the <i>natural</i> doc ID
Modified: lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/index/SegmentReader.java
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/index/SegmentReader.java?rev=1466057&r1=1466056&r2=1466057&view=diff
==============================================================================
--- lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/index/SegmentReader.java (original)
+++ lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/index/SegmentReader.java Tue Apr 9 14:28:52 2013
@@ -192,7 +192,7 @@ public final class SegmentReader extends
/**
* Return the SegmentInfoPerCommit of the segment this reader is reading.
*/
- SegmentInfoPerCommit getSegmentInfo() {
+ public SegmentInfoPerCommit getSegmentInfo() {
return si;
}
Modified: lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/search/IndexSearcher.java
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/search/IndexSearcher.java?rev=1466057&r1=1466056&r2=1466057&view=diff
==============================================================================
--- lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/search/IndexSearcher.java (original)
+++ lucene/dev/branches/branch_4x/lucene/core/src/java/org/apache/lucene/search/IndexSearcher.java Tue Apr 9 14:28:52 2013
@@ -599,7 +599,13 @@ public class IndexSearcher {
// threaded...? the Collector could be sync'd?
// always use single thread:
for (AtomicReaderContext ctx : leaves) { // search each subreader
- collector.setNextReader(ctx);
+ try {
+ collector.setNextReader(ctx);
+ } catch (CollectionTerminatedException e) {
+ // there is no doc of interest in this reader context
+ // continue with the following leaf
+ continue;
+ }
Scorer scorer = weight.scorer(ctx, !collector.acceptsDocsOutOfOrder(), true, ctx.reader().getLiveDocs());
if (scorer != null) {
try {
Modified: lucene/dev/branches/branch_4x/lucene/core/src/test/org/apache/lucene/search/TestEarlyTermination.java
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_4x/lucene/core/src/test/org/apache/lucene/search/TestEarlyTermination.java?rev=1466057&r1=1466056&r2=1466057&view=diff
==============================================================================
--- lucene/dev/branches/branch_4x/lucene/core/src/test/org/apache/lucene/search/TestEarlyTermination.java (original)
+++ lucene/dev/branches/branch_4x/lucene/core/src/test/org/apache/lucene/search/TestEarlyTermination.java Tue Apr 9 14:28:52 2013
@@ -77,7 +77,12 @@ public class TestEarlyTermination extend
@Override
public void setNextReader(AtomicReaderContext context) throws IOException {
- collectionTerminated = false;
+ if (random().nextBoolean()) {
+ collectionTerminated = true;
+ throw new CollectionTerminatedException();
+ } else {
+ collectionTerminated = false;
+ }
}
@Override
Modified: lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/NumericDocValuesSorter.java
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/NumericDocValuesSorter.java?rev=1466057&r1=1466056&r2=1466057&view=diff
==============================================================================
--- lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/NumericDocValuesSorter.java (original)
+++ lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/NumericDocValuesSorter.java Tue Apr 9 14:28:52 2013
@@ -52,4 +52,9 @@ public class NumericDocValuesSorter exte
return sort(reader.maxDoc(), comparator);
}
+ @Override
+ public String getID() {
+ return "DocValues(" + fieldName + ",asc)";
+ }
+
}
Modified: lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/Sorter.java
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/Sorter.java?rev=1466057&r1=1466056&r2=1466057&view=diff
==============================================================================
--- lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/Sorter.java (original)
+++ lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/Sorter.java Tue Apr 9 14:28:52 2013
@@ -26,10 +26,12 @@ import org.apache.lucene.util.SorterTemp
import org.apache.lucene.util.packed.MonotonicAppendingLongBuffer;
/**
- * Sorts documents in a given index by returning a permutation on the docs.
- * Implementations can call {@link #sort(int, DocComparator)} to compute the
- * old-to-new permutation over the given documents and values.
- *
+ * Sorts documents of a given index by returning a permutation on the document
+ * IDs.
+ * <p><b>NOTE</b>: A {@link Sorter} implementation can be easily written from
+ * a {@link DocComparator document comparator} by using the
+ * {@link #sort(int, DocComparator)} helper method. This is especially useful
+ * when documents are directly comparable by their field values.
* @lucene.experimental
*/
public abstract class Sorter {
@@ -48,10 +50,16 @@ public abstract class Sorter {
/** Given the ordinal of a doc ID, return its doc ID in the original index. */
public abstract int newToOld(int docID);
+ /** Return the number of documents in this map. This must be equal to the
+ * {@link AtomicReader#maxDoc() number of documents} of the
+ * {@link AtomicReader} which is sorted. */
+ public abstract int size();
+
}
/** Check consistency of a {@link DocMap}, useful for assertions. */
- static boolean isConsistent(DocMap docMap, int maxDoc) {
+ static boolean isConsistent(DocMap docMap) {
+ final int maxDoc = docMap.size();
for (int i = 0; i < maxDoc; ++i) {
final int newID = docMap.oldToNew(i);
final int oldID = docMap.newToOld(newID);
@@ -73,7 +81,11 @@ public abstract class Sorter {
}
- /** Sorts documents in reverse order. */
+ /** Sorts documents in reverse order.
+ * <b>NOTE</b>: This {@link Sorter} is not idempotent. Sorting an
+ * {@link AtomicReader} once or twice will return two different
+ * {@link AtomicReader} views. This {@link Sorter} should not be used with
+ * {@link SortingMergePolicy}. */
public static final Sorter REVERSE_DOCS = new Sorter() {
@Override
public DocMap sort(final AtomicReader reader) throws IOException {
@@ -87,8 +99,17 @@ public abstract class Sorter {
public int newToOld(int docID) {
return maxDoc - docID - 1;
}
+ @Override
+ public int size() {
+ return maxDoc;
+ }
};
}
+
+ @Override
+ public String getID() {
+ return "ReverseDocs";
+ }
};
private static final class DocValueSorterTemplate extends SorterTemplate {
@@ -179,6 +200,11 @@ public abstract class Sorter {
public int newToOld(int docID) {
return (int) newToOld.get(docID);
}
+
+ @Override
+ public int size() {
+ return maxDoc;
+ }
};
}
@@ -192,8 +218,17 @@ public abstract class Sorter {
* <code>reader</code> is already sorted.
* <p>
* <b>NOTE:</b> deleted documents are expected to appear in the mapping as
- * well, they will however be dropped when the index is actually sorted.
+ * well, they will however be marked as deleted in the sorted view.
*/
public abstract DocMap sort(AtomicReader reader) throws IOException;
-
+
+ /**
+ * Returns the identifier of this {@link Sorter}.
+ * <p>This identifier is similar to {@link Object#hashCode()} and should be
+ * chosen so that two instances of this class that sort documents likewise
+ * will have the same identifier. On the contrary, this identifier should be
+ * different on different {@link Sorter sorters}.
+ */
+ public abstract String getID();
+
}
Modified: lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/SortingAtomicReader.java
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/SortingAtomicReader.java?rev=1466057&r1=1466056&r2=1466057&view=diff
==============================================================================
--- lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/SortingAtomicReader.java (original)
+++ lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/SortingAtomicReader.java Tue Apr 9 14:28:52 2013
@@ -633,7 +633,10 @@ public class SortingAtomicReader extends
// the reader is already sorter
return reader;
}
- assert Sorter.isConsistent(docMap, reader.maxDoc());
+ if (reader.maxDoc() != docMap.size()) {
+ throw new IllegalArgumentException("reader.maxDoc() should be equal to docMap.size(), got" + reader.maxDoc() + " != " + docMap.size());
+ }
+ assert Sorter.isConsistent(docMap);
return new SortingAtomicReader(reader, docMap);
}
Modified: lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/SortingMergePolicy.java
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/SortingMergePolicy.java?rev=1466057&r1=1466056&r2=1466057&view=diff
==============================================================================
--- lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/SortingMergePolicy.java (original)
+++ lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/SortingMergePolicy.java Tue Apr 9 14:28:52 2013
@@ -28,8 +28,10 @@ import org.apache.lucene.index.IndexWrit
import org.apache.lucene.index.MergePolicy;
import org.apache.lucene.index.MergeState;
import org.apache.lucene.index.MultiReader;
+import org.apache.lucene.index.SegmentInfo;
import org.apache.lucene.index.SegmentInfoPerCommit;
import org.apache.lucene.index.SegmentInfos;
+import org.apache.lucene.index.SegmentReader;
import org.apache.lucene.index.SlowCompositeReaderWrapper;
import org.apache.lucene.store.Directory;
import org.apache.lucene.util.Bits;
@@ -39,12 +41,23 @@ import org.apache.lucene.util.packed.Mon
* before merging them. As a consequence, all segments resulting from a merge
* will be sorted while segments resulting from a flush will be in the order
* in which documents have been added.
- * <p>Never use this {@link MergePolicy} if you rely on
+ * <p><b>NOTE</b>: Never use this {@link MergePolicy} if you rely on
* {@link IndexWriter#addDocuments(Iterable, org.apache.lucene.analysis.Analyzer)}
* to have sequentially-assigned doc IDs, this policy will scatter doc IDs.
+ * <p><b>NOTE</b>: This {@link MergePolicy} should only be used with idempotent
+ * {@link Sorter}s so that the order of segments is predictable. For example,
+ * using {@link SortingMergePolicy} with {@link Sorter#REVERSE_DOCS} (which is
+ * not idempotent) will make the order of documents in a segment depend on the
+ * number of times the segment has been merged.
* @lucene.experimental */
public final class SortingMergePolicy extends MergePolicy {
+ /**
+ * Put in the {@link SegmentInfo#getDiagnostics() diagnostics} to denote that
+ * this segment is sorted.
+ */
+ public static final String SORTER_ID_PROP = "sorter";
+
class SortingOneMerge extends OneMerge {
List<AtomicReader> unsortedReaders;
@@ -72,6 +85,13 @@ public final class SortingMergePolicy ex
// a null doc map means that the readers are already sorted
return docMap == null ? unsortedReaders : Collections.singletonList(sortedView);
}
+
+ @Override
+ public void setInfo(SegmentInfoPerCommit info) {
+ Map<String,String> diagnostics = info.info.getDiagnostics();
+ diagnostics.put(SORTER_ID_PROP, sorter.getID());
+ super.setInfo(info);
+ }
private MonotonicAppendingLongBuffer getDeletes(List<AtomicReader> readers) {
MonotonicAppendingLongBuffer deletes = new MonotonicAppendingLongBuffer();
@@ -126,6 +146,18 @@ public final class SortingMergePolicy ex
}
+ /** Returns true if the given reader is sorted by the given sorter. */
+ public static boolean isSorted(AtomicReader reader, Sorter sorter) {
+ if (reader instanceof SegmentReader) {
+ final SegmentReader segReader = (SegmentReader) reader;
+ final Map<String, String> diagnostics = segReader.getSegmentInfo().info.getDiagnostics();
+ if (diagnostics != null && sorter.getID().equals(diagnostics.get(SORTER_ID_PROP))) {
+ return true;
+ }
+ }
+ return false;
+ }
+
private MergeSpecification sortedMergeSpecification(MergeSpecification specification) {
if (specification == null) {
return null;
Modified: lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/package.html
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/package.html?rev=1466057&r1=1466056&r2=1466057&view=diff
==============================================================================
--- lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/package.html (original)
+++ lucene/dev/branches/branch_4x/lucene/misc/src/java/org/apache/lucene/index/sorter/package.html Tue Apr 9 14:28:52 2013
@@ -17,11 +17,25 @@
-->
<html>
<body>
-Provides index sorting capablities. The application can use one of the
-pre-existing Sorter implementations, e.g. to sort by a numeric
-DocValues or reverse the order of the documents. Additionally, the
-application can implement a Sorter which returns a permutation on
-a source Directory's document IDs, to sort the input documents by additional
-values.
+<p>Provides index sorting capablities. The application can use one of the
+pre-existing Sorter implementations, e.g. to sort by a
+{@link org.apache.lucene.index.sorter.NumericDocValuesSorter}
+or {@link org.apache.lucene.index.sorter.Sorter#REVERSE_DOCS reverse} the order
+of the documents. Additionally, the application can implement a custom
+{@link org.apache.lucene.index.sorter.Sorter} which returns a permutation on
+a source {@link org.apache.lucene.index.AtomicReader}'s document IDs, to sort
+the input documents by additional criteria.
+
+<p>{@link org.apache.lucene.index.sorter.SortingMergePolicy} can be used to
+make Lucene sort segments before merging them. This will ensure that every
+segment resulting from a merge will be sorted according to the provided
+{@link org.apache.lucene.index.sorter.Sorter}. This however makes merging and
+thus indexing slower.
+
+<p>Sorted segments allow for early query termination when the sort order
+matches index order. This makes query execution faster since not all documents
+need to be visited. Please note that this is an expert feature and should not
+be used without a deep understanding of Lucene merging and document collection.
+
</body>
</html>
Modified: lucene/dev/branches/branch_4x/lucene/misc/src/test/org/apache/lucene/index/sorter/SortingAtomicReaderTest.java
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_4x/lucene/misc/src/test/org/apache/lucene/index/sorter/SortingAtomicReaderTest.java?rev=1466057&r1=1466056&r2=1466057&view=diff
==============================================================================
--- lucene/dev/branches/branch_4x/lucene/misc/src/test/org/apache/lucene/index/sorter/SortingAtomicReaderTest.java (original)
+++ lucene/dev/branches/branch_4x/lucene/misc/src/test/org/apache/lucene/index/sorter/SortingAtomicReaderTest.java Tue Apr 9 14:28:52 2013
@@ -61,6 +61,10 @@ public class SortingAtomicReaderTest ext
public Sorter.DocMap sort(AtomicReader reader) throws IOException {
return docMap;
}
+ @Override
+ public String getID() {
+ return ID_FIELD;
+ }
});
if (VERBOSE) {
Modified: lucene/dev/branches/branch_4x/lucene/misc/src/test/org/apache/lucene/index/sorter/TestSortingMergePolicy.java
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_4x/lucene/misc/src/test/org/apache/lucene/index/sorter/TestSortingMergePolicy.java?rev=1466057&r1=1466056&r2=1466057&view=diff
==============================================================================
--- lucene/dev/branches/branch_4x/lucene/misc/src/test/org/apache/lucene/index/sorter/TestSortingMergePolicy.java (original)
+++ lucene/dev/branches/branch_4x/lucene/misc/src/test/org/apache/lucene/index/sorter/TestSortingMergePolicy.java Tue Apr 9 14:28:52 2013
@@ -70,7 +70,7 @@ public class TestSortingMergePolicy exte
return doc;
}
- private static MergePolicy newSortingMergePolicy(Sorter sorter) {
+ static MergePolicy newSortingMergePolicy(Sorter sorter) {
// create a MP with a low merge factor so that many merges happen
MergePolicy mp;
if (random().nextBoolean()) {