Update Apache Lucene to 9.12.0 (opensearch-project#15333)

* Update Apache Lucene to 9.12.0

Signed-off-by: Andriy Redko <[email protected]>

* change to IOContext READONCE in locations where the file is not expected to be read multiple times.

Signed-off-by: Marc Handalian <[email protected]>

* Use READ IOContext for all non Segment* files when copying node-node

Signed-off-by: Marc Handalian <[email protected]>

* Fixing more test failures

Signed-off-by: Andriy Redko <[email protected]>

* Move Composite912Codec under org.opensearch.index.codec.composite.composite912 package

Signed-off-by: Andriy Redko <[email protected]>

---------

Signed-off-by: Andriy Redko <[email protected]>
Signed-off-by: Marc Handalian <[email protected]>
Co-authored-by: Marc Handalian <[email protected]>
(cherry picked from commit 146b0f7)
Signed-off-by: Andriy Redko <[email protected]>

Author: reta

CHANGELOG.md

@@ -1,5 +1,5 @@
 opensearch        = 2.18.0
-lucene            = 9.11.1
+lucene            = 9.12.0
 
 bundled_jdk_vendor = adoptium
 bundled_jdk = 21.0.4+7

buildSrc/version.properties

@@ -0,0 +1 @@
+fdb055d569bb20bfce9618fe2b01c29bab7f290c

libs/core/licenses/lucene-core-9.11.1.jar.sha1

@@ -141,7 +141,7 @@ public class Version implements Comparable<Version>, ToXContentFragment {
     public static final Version V_2_17_0 = new Version(2170099, org.apache.lucene.util.Version.LUCENE_9_11_1);
     public static final Version V_2_17_1 = new Version(2170199, org.apache.lucene.util.Version.LUCENE_9_11_1);
     public static final Version V_2_17_2 = new Version(2170299, org.apache.lucene.util.Version.LUCENE_9_11_1);
-    public static final Version V_2_18_0 = new Version(2180099, org.apache.lucene.util.Version.LUCENE_9_11_1);
+    public static final Version V_2_18_0 = new Version(2180099, org.apache.lucene.util.Version.LUCENE_9_12_0);
     public static final Version CURRENT = V_2_18_0;
 
     public static Version fromId(int id) {

libs/core/licenses/lucene-core-9.12.0.jar.sha1

@@ -0,0 +1 @@
+5ba843374a0aab3dfe0b11cb28b251844d85bf5b

libs/core/src/main/java/org/opensearch/Version.java

@@ -0,0 +1 @@
+a9232b6a4882979118d3281b98dfdb6e0e1cb5ca

modules/lang-expression/licenses/lucene-expressions-9.11.1.jar.sha1

@@ -0,0 +1 @@
+a3a6950ffc22e76a082e1b3cefb022b9f7870d29

modules/lang-expression/licenses/lucene-expressions-9.12.0.jar.sha1

@@ -0,0 +1 @@
+e71f85b72ed3939039ba8897b28b065dd11918b9

plugins/analysis-icu/licenses/lucene-analysis-icu-9.11.1.jar.sha1

@@ -0,0 +1 @@
+6baa3ae7ab20d6e644cf0bedb271c50a44c0e259

plugins/analysis-icu/licenses/lucene-analysis-icu-9.12.0.jar.sha1

@@ -0,0 +1 @@
+f183e1e8b1eaaa4dec444774a285bb8b66518522

plugins/analysis-kuromoji/licenses/lucene-analysis-kuromoji-9.11.1.jar.sha1

@@ -0,0 +1 @@
+b81a609934e65d12ab9d2d84bc2ea6f56a360e57

plugins/analysis-kuromoji/licenses/lucene-analysis-kuromoji-9.12.0.jar.sha1

@@ -0,0 +1 @@
+bec069f286b45f20b743c81e84202369cd0467e7

plugins/analysis-nori/licenses/lucene-analysis-nori-9.11.1.jar.sha1

@@ -0,0 +1 @@
+4c2503cfaba37249e20ea877555cb52ee89d1ae1

plugins/analysis-nori/licenses/lucene-analysis-nori-9.12.0.jar.sha1

@@ -0,0 +1 @@
+68fe98c94e9644a584ea1bf525e68d9406fc61ec

plugins/analysis-phonetic/licenses/lucene-analysis-phonetic-9.11.1.jar.sha1

@@ -0,0 +1 @@
+fdb055d569bb20bfce9618fe2b01c29bab7f290c

plugins/analysis-phonetic/licenses/lucene-analysis-phonetic-9.12.0.jar.sha1

@@ -0,0 +1 @@
+ccf99f8db57aa97b2c1f95c5cc2a11156a043921

plugins/analysis-smartcn/licenses/lucene-analysis-smartcn-9.11.1.jar.sha1

@@ -0,0 +1 @@
+e93429f66fbcd3b58d81f01223d6ce5688047296

plugins/analysis-smartcn/licenses/lucene-analysis-smartcn-9.12.0.jar.sha1

@@ -0,0 +1 @@
+14c802d6955eaf11107375a2ada8fe8ec53b3e01

plugins/analysis-stempel/licenses/lucene-analysis-stempel-9.11.1.jar.sha1

@@ -0,0 +1 @@
+ffe090339540876b40df792aee51a42af6b3f37f

plugins/analysis-stempel/licenses/lucene-analysis-stempel-9.12.0.jar.sha1

@@ -0,0 +1 @@
+ad17704ee90eb926b6d3105f7027485cdadbecd9

plugins/analysis-ukrainian/licenses/lucene-analysis-morfologik-9.11.1.jar.sha1

@@ -0,0 +1 @@
+3ac2a62b0b55c5725bb65f0c5454f9f8a401cf43

plugins/analysis-ukrainian/licenses/lucene-analysis-morfologik-9.12.0.jar.sha1

@@ -0,0 +1 @@
+55959399373876f4c184944315458dc6b88fbd81

server/licenses/lucene-analysis-common-9.11.1.jar.sha1

@@ -0,0 +1 @@
+f65882536d681c11a1cbc920e5679201101e3603

server/licenses/lucene-analysis-common-9.12.0.jar.sha1

@@ -0,0 +1 @@
+9d00cc7cc2279822ef6740f0677cafacfb439fa8

server/licenses/lucene-backward-codecs-9.11.1.jar.sha1

@@ -0,0 +1 @@
+e3092632ca1d4427d3ebb2c866ac89d90f5b61ec

server/licenses/lucene-backward-codecs-9.12.0.jar.sha1

@@ -0,0 +1 @@
+e1c6636499317ebe498f3490a1ec8b86b8a363dd

server/licenses/lucene-core-9.11.1.jar.sha1

@@ -110,7 +110,7 @@ public void testConcurrentSearchTaskTracking() {
             Map<Long, List<ThreadResourceInfo>> threadStats = getThreadStats(SearchAction.NAME + "[*]", taskInfo.getTaskId());
             // Concurrent search forks each slice of 5 segments to different thread (see please
             // https://github.com/apache/lucene/issues/12498)
-            assertEquals((int) Math.ceil(getSegmentCount(INDEX_NAME) / 5.0) + 1, threadStats.size());
+            assertEquals((int) Math.ceil(getSegmentCount(INDEX_NAME) / 5.0), threadStats.size());
 
             // assert that all task descriptions have non-zero length
             MatcherAssert.assertThat(taskInfo.getDescription().length(), greaterThan(0));

server/licenses/lucene-core-9.12.0.jar.sha1

@@ -58,6 +58,7 @@
 import java.util.Collections;
 import java.util.List;
 import java.util.concurrent.CountDownLatch;
+import java.util.concurrent.TimeUnit;
 import java.util.concurrent.atomic.AtomicBoolean;
 
 import static org.opensearch.indices.recovery.RecoverySettings.INDICES_RECOVERY_CHUNK_SIZE_SETTING;
@@ -184,7 +185,7 @@ public void testCancelRecoveryAndResume() throws Exception {
             )
             .get();
 
-        latch.await();
+        latch.await(5, TimeUnit.MINUTES);
 
         // at this point we got some truncated left overs on the replica on the unlucky node
         // now we are allowing the recovery to allocate again and finish to see if we wipe the truncated files

server/licenses/lucene-grouping-9.11.1.jar.sha1

@@ -70,6 +70,7 @@ public LegacyBM25Similarity(float k1, float b) {
      *     within the range {@code [0..1]}
      */
     public LegacyBM25Similarity(float k1, float b, boolean discountOverlaps) {
+        super(discountOverlaps);
         this.bm25Similarity = new BM25Similarity(k1, b, discountOverlaps);
     }
 
@@ -101,15 +102,6 @@ public final float getB() {
         return bm25Similarity.getB();
     }
 
-    /**
-     * Returns true if overlap tokens are discounted from the document's length.
-     *
-     * @see #LegacyBM25Similarity(float, float, boolean)
-     */
-    public boolean getDiscountOverlaps() {
-        return bm25Similarity.getDiscountOverlaps();
-    }
-
     @Override
     public String toString() {
         return bm25Similarity.toString();

server/licenses/lucene-grouping-9.12.0.jar.sha1

@@ -87,15 +87,7 @@ public int prevSetBit(int index) {
 
     @Override
     public int nextSetBit(int index) {
-        assert index >= 0 && index < length : "index=" + index + " numBits=" + length();
-        int next = first.nextSetBit(index);
-        while (next != DocIdSetIterator.NO_MORE_DOCS && second.get(next) == false) {
-            if (next == length() - 1) {
-                return DocIdSetIterator.NO_MORE_DOCS;
-            }
-            next = first.nextSetBit(next + 1);
-        }
-        return next;
+        return nextSetBit(index, length() - 1);
     }
 
     @Override
@@ -132,4 +124,20 @@ public void clear(int startIndex, int endIndex) {
     public boolean getAndSet(int i) {
         throw new UnsupportedOperationException("not implemented");
     }
+
+    @Override
+    public int nextSetBit(int start, int end) {
+        assert start >= 0 && start < length() : "start=" + start + " numBits=" + length();
+        assert end >= start && end < length() : "end=" + end + " numBits=" + length();
+
+        int next = first.nextSetBit(start);
+        while (next != DocIdSetIterator.NO_MORE_DOCS && second.get(next) == false) {
+            if (next >= end) {
+                return DocIdSetIterator.NO_MORE_DOCS;
+            }
+            next = first.nextSetBit(next + 1);
+        }
+        return next;
+
+    }
 }

server/licenses/lucene-highlighter-9.11.1.jar.sha1

@@ -113,7 +113,7 @@
  * @opensearch.internal
  */
 public class Lucene {
-    public static final String LATEST_CODEC = "Lucene99";
+    public static final String LATEST_CODEC = "Lucene912";
 
     public static final String SOFT_DELETES_FIELD = "__soft_deletes";
 
@@ -287,7 +287,7 @@ public static void checkSegmentInfoIntegrity(final Directory directory) throws I
 
             @Override
             protected Object doBody(String segmentFileName) throws IOException {
-                try (IndexInput input = directory.openInput(segmentFileName, IOContext.READ)) {
+                try (IndexInput input = directory.openInput(segmentFileName, IOContext.READONCE)) {
                     CodecUtil.checksumEntireFile(input);
                 }
                 return null;

server/licenses/lucene-highlighter-9.12.0.jar.sha1

@@ -34,8 +34,8 @@
 
 import org.apache.logging.log4j.Logger;
 import org.apache.lucene.codecs.Codec;
-import org.apache.lucene.codecs.lucene99.Lucene99Codec;
-import org.apache.lucene.codecs.lucene99.Lucene99Codec.Mode;
+import org.apache.lucene.codecs.lucene912.Lucene912Codec;
+import org.apache.lucene.codecs.lucene912.Lucene912Codec.Mode;
 import org.opensearch.common.Nullable;
 import org.opensearch.common.collect.MapBuilder;
 import org.opensearch.index.IndexSettings;
@@ -70,10 +70,10 @@ public CodecService(@Nullable MapperService mapperService, IndexSettings indexSe
         final MapBuilder<String, Codec> codecs = MapBuilder.<String, Codec>newMapBuilder();
         assert null != indexSettings;
         if (mapperService == null) {
-            codecs.put(DEFAULT_CODEC, new Lucene99Codec());
-            codecs.put(LZ4, new Lucene99Codec());
-            codecs.put(BEST_COMPRESSION_CODEC, new Lucene99Codec(Mode.BEST_COMPRESSION));
-            codecs.put(ZLIB, new Lucene99Codec(Mode.BEST_COMPRESSION));
+            codecs.put(DEFAULT_CODEC, new Lucene912Codec());
+            codecs.put(LZ4, new Lucene912Codec());
+            codecs.put(BEST_COMPRESSION_CODEC, new Lucene912Codec(Mode.BEST_COMPRESSION));
+            codecs.put(ZLIB, new Lucene912Codec(Mode.BEST_COMPRESSION));
         } else {
             // CompositeCodec still delegates to PerFieldMappingPostingFormatCodec
             // We can still support all the compression codecs when composite index is present
@@ -97,8 +97,8 @@ public CodecService(@Nullable MapperService mapperService, IndexSettings indexSe
     public CodecService(@Nullable MapperService mapperService, Logger logger) {
         final MapBuilder<String, Codec> codecs = MapBuilder.<String, Codec>newMapBuilder();
         if (mapperService == null) {
-            codecs.put(DEFAULT_CODEC, new Lucene99Codec());
-            codecs.put(BEST_COMPRESSION_CODEC, new Lucene99Codec(Mode.BEST_COMPRESSION));
+            codecs.put(DEFAULT_CODEC, new Lucene912Codec());
+            codecs.put(BEST_COMPRESSION_CODEC, new Lucene912Codec(Mode.BEST_COMPRESSION));
         } else {
             IndexSettings indexSettings = mapperService.getIndexSettings();
             codecs.put(DEFAULT_CODEC, new PerFieldMappingPostingFormatCodec(Mode.BEST_SPEED, mapperService, logger));

server/licenses/lucene-join-9.11.1.jar.sha1

@@ -37,7 +37,7 @@
 import org.apache.lucene.codecs.DocValuesFormat;
 import org.apache.lucene.codecs.PostingsFormat;
 import org.apache.lucene.codecs.lucene90.Lucene90DocValuesFormat;
-import org.apache.lucene.codecs.lucene99.Lucene99Codec;
+import org.apache.lucene.codecs.lucene912.Lucene912Codec;
 import org.opensearch.common.lucene.Lucene;
 import org.opensearch.index.codec.fuzzy.FuzzyFilterPostingsFormat;
 import org.opensearch.index.codec.fuzzy.FuzzySetFactory;
@@ -59,7 +59,7 @@
  *
  * @opensearch.internal
  */
-public class PerFieldMappingPostingFormatCodec extends Lucene99Codec {
+public class PerFieldMappingPostingFormatCodec extends Lucene912Codec {
     private final Logger logger;
     private final MapperService mapperService;
     private final DocValuesFormat dvFormat = new Lucene90DocValuesFormat();

server/licenses/lucene-join-9.12.0.jar.sha1

@@ -10,8 +10,9 @@
 
 import org.apache.logging.log4j.Logger;
 import org.apache.lucene.codecs.Codec;
-import org.apache.lucene.codecs.lucene99.Lucene99Codec;
+import org.apache.lucene.codecs.lucene912.Lucene912Codec;
 import org.opensearch.common.annotation.ExperimentalApi;
+import org.opensearch.index.codec.composite.composite912.Composite912Codec;
 import org.opensearch.index.codec.composite.composite99.Composite99Codec;
 import org.opensearch.index.mapper.MapperService;
 
@@ -38,10 +39,10 @@ public CompositeCodecFactory() {}
 
     public Map<String, Codec> getCompositeIndexCodecs(MapperService mapperService, Logger logger) {
         Map<String, Codec> codecs = new HashMap<>();
-        codecs.put(DEFAULT_CODEC, new Composite99Codec(Lucene99Codec.Mode.BEST_SPEED, mapperService, logger));
-        codecs.put(LZ4, new Composite99Codec(Lucene99Codec.Mode.BEST_SPEED, mapperService, logger));
-        codecs.put(BEST_COMPRESSION_CODEC, new Composite99Codec(Lucene99Codec.Mode.BEST_COMPRESSION, mapperService, logger));
-        codecs.put(ZLIB, new Composite99Codec(Lucene99Codec.Mode.BEST_COMPRESSION, mapperService, logger));
+        codecs.put(DEFAULT_CODEC, new Composite912Codec(Lucene912Codec.Mode.BEST_SPEED, mapperService, logger));
+        codecs.put(LZ4, new Composite912Codec(Lucene912Codec.Mode.BEST_SPEED, mapperService, logger));
+        codecs.put(BEST_COMPRESSION_CODEC, new Composite912Codec(Lucene912Codec.Mode.BEST_COMPRESSION, mapperService, logger));
+        codecs.put(ZLIB, new Composite912Codec(Lucene912Codec.Mode.BEST_COMPRESSION, mapperService, logger));
         return codecs;
     }
 }