LUCENE-7257: PointValues aggregated stats fail if the provided field does not have points on one of the leaves.

This commit is contained in:
Adrien Grand 2016-04-27 11:40:32 +02:00
parent 9ce830d8f2
commit c7cdf2832b
3 changed files with 78 additions and 14 deletions

View File

@ -104,6 +104,10 @@ Bug Fixes
* LUCENE-7232: Fixed InetAddressPoint.newPrefixQuery, which was generating an
incorrect query when the prefix length was not a multiple of 8. (Adrien Grand)
* LUCENE-7257: Fixed PointValues#size(IndexReader, String), docCount,
minPackedValue and maxPackedValue to skip leaves that do not have points
rather than raising an IllegalStateException. (Adrien Grand)
Documentation
* LUCENE-7223: Improve XXXPoint javadocs to make it clear that you

View File

@ -87,43 +87,51 @@ public abstract class PointValues {
public static final int MAX_DIMENSIONS = BKDWriter.MAX_DIMS;
/** Return the cumulated number of points across all leaves of the given
* {@link IndexReader}.
* {@link IndexReader}. Leaves that do not have points for the given field
* are ignored.
* @see PointValues#size(String) */
public static long size(IndexReader reader, String field) throws IOException {
long size = 0;
for (LeafReaderContext ctx : reader.leaves()) {
PointValues values = ctx.reader().getPointValues();
if (values != null) {
size += values.size(field);
FieldInfo info = ctx.reader().getFieldInfos().fieldInfo(field);
if (info == null || info.getPointDimensionCount() == 0) {
continue;
}
PointValues values = ctx.reader().getPointValues();
size += values.size(field);
}
return size;
}
/** Return the cumulated number of docs that have points across all leaves
* of the given {@link IndexReader}.
* of the given {@link IndexReader}. Leaves that do not have points for the
* given field are ignored.
* @see PointValues#getDocCount(String) */
public static int getDocCount(IndexReader reader, String field) throws IOException {
int count = 0;
for (LeafReaderContext ctx : reader.leaves()) {
PointValues values = ctx.reader().getPointValues();
if (values != null) {
count += values.getDocCount(field);
FieldInfo info = ctx.reader().getFieldInfos().fieldInfo(field);
if (info == null || info.getPointDimensionCount() == 0) {
continue;
}
PointValues values = ctx.reader().getPointValues();
count += values.getDocCount(field);
}
return count;
}
/** Return the minimum packed values across all leaves of the given
* {@link IndexReader}.
* {@link IndexReader}. Leaves that do not have points for the given field
* are ignored.
* @see PointValues#getMinPackedValue(String) */
public static byte[] getMinPackedValue(IndexReader reader, String field) throws IOException {
byte[] minValue = null;
for (LeafReaderContext ctx : reader.leaves()) {
PointValues values = ctx.reader().getPointValues();
if (values == null) {
FieldInfo info = ctx.reader().getFieldInfos().fieldInfo(field);
if (info == null || info.getPointDimensionCount() == 0) {
continue;
}
PointValues values = ctx.reader().getPointValues();
byte[] leafMinValue = values.getMinPackedValue(field);
if (leafMinValue == null) {
continue;
@ -145,15 +153,17 @@ public abstract class PointValues {
}
/** Return the maximum packed values across all leaves of the given
* {@link IndexReader}.
* {@link IndexReader}. Leaves that do not have points for the given field
* are ignored.
* @see PointValues#getMaxPackedValue(String) */
public static byte[] getMaxPackedValue(IndexReader reader, String field) throws IOException {
byte[] maxValue = null;
for (LeafReaderContext ctx : reader.leaves()) {
PointValues values = ctx.reader().getPointValues();
if (values == null) {
FieldInfo info = ctx.reader().getFieldInfos().fieldInfo(field);
if (info == null || info.getPointDimensionCount() == 0) {
continue;
}
PointValues values = ctx.reader().getPointValues();
byte[] leafMaxValue = values.getMaxPackedValue(field);
if (leafMaxValue == null) {
continue;

View File

@ -30,6 +30,7 @@ import org.apache.lucene.document.FloatPoint;
import org.apache.lucene.document.IntPoint;
import org.apache.lucene.document.LongPoint;
import org.apache.lucene.document.StringField;
import org.apache.lucene.document.Field.Store;
import org.apache.lucene.index.PointValues.IntersectVisitor;
import org.apache.lucene.index.PointValues.Relation;
import org.apache.lucene.index.PointValues;
@ -652,6 +653,55 @@ public class TestPointValues extends LuceneTestCase {
dir.close();
}
public void testMergedStatsEmptyReader() throws IOException {
IndexReader reader = new MultiReader();
assertNull(PointValues.getMinPackedValue(reader, "field"));
assertNull(PointValues.getMaxPackedValue(reader, "field"));
assertEquals(0, PointValues.getDocCount(reader, "field"));
assertEquals(0, PointValues.size(reader, "field"));
}
public void testMergedStatsOneSegmentWithoutPoints() throws IOException {
Directory dir = new RAMDirectory();
IndexWriter w = new IndexWriter(dir, new IndexWriterConfig(null).setMergePolicy(NoMergePolicy.INSTANCE));
w.addDocument(new Document());
DirectoryReader.open(w).close();
Document doc = new Document();
doc.add(new IntPoint("field", Integer.MIN_VALUE));
w.addDocument(doc);
IndexReader reader = DirectoryReader.open(w);
assertArrayEquals(new byte[4], PointValues.getMinPackedValue(reader, "field"));
assertArrayEquals(new byte[4], PointValues.getMaxPackedValue(reader, "field"));
assertEquals(1, PointValues.getDocCount(reader, "field"));
assertEquals(1, PointValues.size(reader, "field"));
assertNull(PointValues.getMinPackedValue(reader, "field2"));
assertNull(PointValues.getMaxPackedValue(reader, "field2"));
assertEquals(0, PointValues.getDocCount(reader, "field2"));
assertEquals(0, PointValues.size(reader, "field2"));
}
public void testMergedStatsAllPointsDeleted() throws IOException {
Directory dir = new RAMDirectory();
IndexWriter w = new IndexWriter(dir, new IndexWriterConfig(null));
w.addDocument(new Document());
Document doc = new Document();
doc.add(new IntPoint("field", Integer.MIN_VALUE));
doc.add(new StringField("delete", "yes", Store.NO));
w.addDocument(doc);
w.forceMerge(1);
w.deleteDocuments(new Term("delete", "yes"));
w.addDocument(new Document());
w.forceMerge(1);
IndexReader reader = DirectoryReader.open(w);
assertNull(PointValues.getMinPackedValue(reader, "field"));
assertNull(PointValues.getMaxPackedValue(reader, "field"));
assertEquals(0, PointValues.getDocCount(reader, "field"));
assertEquals(0, PointValues.size(reader, "field"));
}
public void testMergedStats() throws IOException {
final int iters = atLeast(3);
for (int iter = 0; iter < iters; ++iter) {