You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@lucene.apache.org by mi...@apache.org on 2013/10/29 18:33:05 UTC
svn commit: r1536826 - in /lucene/dev/trunk/lucene/suggest/src:
java/org/apache/lucene/search/suggest/DocumentExpressionDictionary.java
test/org/apache/lucene/search/suggest/DocumentExpressionDictionaryTest.java
Author: mikemccand
Date: Tue Oct 29 17:33:04 2013
New Revision: 1536826
URL: http://svn.apache.org/r1536826
Log:
LUCENE-5306: DocumentExpressionDictionary now accepts composite readers
Modified:
lucene/dev/trunk/lucene/suggest/src/java/org/apache/lucene/search/suggest/DocumentExpressionDictionary.java
lucene/dev/trunk/lucene/suggest/src/test/org/apache/lucene/search/suggest/DocumentExpressionDictionaryTest.java
Modified: lucene/dev/trunk/lucene/suggest/src/java/org/apache/lucene/search/suggest/DocumentExpressionDictionary.java
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/suggest/src/java/org/apache/lucene/search/suggest/DocumentExpressionDictionary.java?rev=1536826&r1=1536825&r2=1536826&view=diff
==============================================================================
--- lucene/dev/trunk/lucene/suggest/src/java/org/apache/lucene/search/suggest/DocumentExpressionDictionary.java (original)
+++ lucene/dev/trunk/lucene/suggest/src/java/org/apache/lucene/search/suggest/DocumentExpressionDictionary.java Tue Oct 29 17:33:04 2013
@@ -28,8 +28,8 @@ import org.apache.lucene.expressions.Exp
import org.apache.lucene.expressions.SimpleBindings;
import org.apache.lucene.expressions.js.JavascriptCompiler;
import org.apache.lucene.index.AtomicReaderContext;
-import org.apache.lucene.index.CompositeReader; // javadocs
import org.apache.lucene.index.IndexReader;
+import org.apache.lucene.index.ReaderUtil;
import org.apache.lucene.queries.function.FunctionValues;
import org.apache.lucene.queries.function.ValueSource;
import org.apache.lucene.search.SortField;
@@ -49,9 +49,6 @@ import org.apache.lucene.util.BytesRefIt
* The term and (optionally) payload fields supplied
* are required for ALL documents and has to be stored
* </li>
- * <li>
- * {@link CompositeReader} is not supported.
- * </li>
* </ul>
*/
public class DocumentExpressionDictionary extends DocumentDictionary {
@@ -100,21 +97,41 @@ public class DocumentExpressionDictionar
final class DocumentExpressionInputIterator extends DocumentDictionary.DocumentInputIterator {
- private FunctionValues weightValues;
+ private FunctionValues currentWeightValues;
+ private int currentLeafIndex = 0;
+ private final List<AtomicReaderContext> leaves;
+
+ private final int[] starts;
public DocumentExpressionInputIterator(boolean hasPayloads)
throws IOException {
super(hasPayloads);
- List<AtomicReaderContext> leaves = reader.leaves();
- if (leaves.size() > 1) {
- throw new IllegalArgumentException("CompositeReader is not supported");
+ leaves = reader.leaves();
+ if (leaves.size() == 0) {
+ throw new IllegalArgumentException("Reader has to have at least one leaf");
}
- weightValues = weightsValueSource.getValues(new HashMap<String, Object>(), leaves.get(0));
+ starts = new int[leaves.size() + 1];
+ for (int i = 0; i < leaves.size(); i++) {
+ starts[i] = leaves.get(i).docBase;
+ }
+ starts[leaves.size()] = reader.maxDoc();
+
+ currentLeafIndex = 0;
+ currentWeightValues = weightsValueSource.getValues(new HashMap<String, Object>(), leaves.get(currentLeafIndex));
}
@Override
protected long getWeight(int docId) {
- return weightValues.longVal(docId);
+ int subIndex = ReaderUtil.subIndex(docId, starts);
+ if (subIndex != currentLeafIndex) {
+ currentLeafIndex = subIndex;
+ try {
+ currentWeightValues = weightsValueSource.getValues(new HashMap<String, Object>(), leaves.get(currentLeafIndex));
+ } catch (IOException e) {
+ throw new RuntimeException();
+ }
+ }
+ return currentWeightValues.longVal(docId - starts[subIndex]);
}
}
Modified: lucene/dev/trunk/lucene/suggest/src/test/org/apache/lucene/search/suggest/DocumentExpressionDictionaryTest.java
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/suggest/src/test/org/apache/lucene/search/suggest/DocumentExpressionDictionaryTest.java?rev=1536826&r1=1536825&r2=1536826&view=diff
==============================================================================
--- lucene/dev/trunk/lucene/suggest/src/test/org/apache/lucene/search/suggest/DocumentExpressionDictionaryTest.java (original)
+++ lucene/dev/trunk/lucene/suggest/src/test/org/apache/lucene/search/suggest/DocumentExpressionDictionaryTest.java Tue Oct 29 17:33:04 2013
@@ -36,7 +36,6 @@ import org.apache.lucene.index.Directory
import org.apache.lucene.index.IndexReader;
import org.apache.lucene.index.IndexWriterConfig;
import org.apache.lucene.index.RandomIndexWriter;
-import org.apache.lucene.index.SlowCompositeReaderWrapper;
import org.apache.lucene.index.Term;
import org.apache.lucene.search.SortField;
import org.apache.lucene.search.spell.Dictionary;
@@ -78,16 +77,14 @@ public class DocumentExpressionDictionar
IndexWriterConfig iwc = newIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(random()));
iwc.setMergePolicy(newLogMergePolicy());
RandomIndexWriter writer = new RandomIndexWriter(random(), dir, iwc);
- Map<String, Document> docs = generateIndexDocuments(10);
+ Map<String, Document> docs = generateIndexDocuments(atLeast(10));
for(Document doc: docs.values()) {
writer.addDocument(doc);
}
writer.commit();
writer.close();
- // TODO: once we fix DocumentExpressionDictionary to
- // accept readers with more than one segment, we can
- // remove this wrapping:
- IndexReader ir = SlowCompositeReaderWrapper.wrap(DirectoryReader.open(dir));
+
+ IndexReader ir = DirectoryReader.open(dir);
Set<SortField> sortFields = new HashSet<SortField>();
sortFields.add(new SortField(WEIGHT_FIELD_NAME_1, SortField.Type.LONG));
sortFields.add(new SortField(WEIGHT_FIELD_NAME_2, SortField.Type.LONG));
@@ -115,16 +112,14 @@ public class DocumentExpressionDictionar
IndexWriterConfig iwc = newIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(random()));
iwc.setMergePolicy(newLogMergePolicy());
RandomIndexWriter writer = new RandomIndexWriter(random(), dir, iwc);
- Map<String, Document> docs = generateIndexDocuments(10);
+ Map<String, Document> docs = generateIndexDocuments(atLeast(10));
for(Document doc: docs.values()) {
writer.addDocument(doc);
}
writer.commit();
writer.close();
- // TODO: once we fix DocumentExpressionDictionary to
- // accept readers with more than one segment, we can
- // remove this wrapping:
- IndexReader ir = SlowCompositeReaderWrapper.wrap(DirectoryReader.open(dir));
+
+ IndexReader ir = DirectoryReader.open(dir);
Set<SortField> sortFields = new HashSet<SortField>();
sortFields.add(new SortField(WEIGHT_FIELD_NAME_1, SortField.Type.LONG));
sortFields.add(new SortField(WEIGHT_FIELD_NAME_2, SortField.Type.LONG));
@@ -152,7 +147,7 @@ public class DocumentExpressionDictionar
IndexWriterConfig iwc = newIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(random()));
iwc.setMergePolicy(newLogMergePolicy());
RandomIndexWriter writer = new RandomIndexWriter(random(), dir, iwc);
- Map<String, Document> docs = generateIndexDocuments(10);
+ Map<String, Document> docs = generateIndexDocuments(atLeast(10));
Random rand = random();
List<String> termsToDel = new ArrayList<>();
for(Document doc : docs.values()) {
@@ -178,10 +173,7 @@ public class DocumentExpressionDictionar
assertTrue(null!=docs.remove(termToDel));
}
- // TODO: once we fix DocumentExpressionDictionary to
- // accept readers with more than one segment, we can
- // remove this wrapping:
- IndexReader ir = SlowCompositeReaderWrapper.wrap(DirectoryReader.open(dir));
+ IndexReader ir = DirectoryReader.open(dir);
assertEquals(ir.numDocs(), docs.size());
Set<SortField> sortFields = new HashSet<SortField>();
sortFields.add(new SortField(WEIGHT_FIELD_NAME_1, SortField.Type.LONG));