Speed up IndexedDISI Sparse #AdvanceExactWithinBlock for tiny step ad…

…vance (#12324)
apache · Jun 13, 2023 · 30eba6d · 30eba6d
1 parent c8e05c8
commit 30eba6d
Show file tree

Hide file tree

Showing 2 changed files with 14 additions and 4 deletions.
diff --git a/lucene/CHANGES.txt b/lucene/CHANGES.txt
@@ -173,6 +173,8 @@ Improvements
 Optimizations
 ---------------------
 
+* GITHUB#12324: Speed up sparse block advanceExact with tiny step in IndexedDISI. (Guo Feng)
+
 * GITHUB#12270 Don't generate stacktrace in CollectionTerminatedException. (Armin Braun)
 
 * GITHUB#12160: Concurrent rewrite for AbstractKnnVectorQuery. (Kaival Parikh)

diff --git a/lucene/core/src/java/org/apache/lucene/codecs/lucene90/IndexedDISI.java b/lucene/core/src/java/org/apache/lucene/codecs/lucene90/IndexedDISI.java
@@ -110,12 +110,12 @@ public final class IndexedDISI extends DocIdSetIterator {
   private static void flush(
       int block, FixedBitSet buffer, int cardinality, byte denseRankPower, IndexOutput out)
       throws IOException {
-    assert block >= 0 && block < 65536;
+    assert block >= 0 && block < BLOCK_SIZE;
     out.writeShort((short) block);
-    assert cardinality > 0 && cardinality <= 65536;
+    assert cardinality > 0 && cardinality <= BLOCK_SIZE;
     out.writeShort((short) (cardinality - 1));
     if (cardinality > MAX_ARRAY_LENGTH) {
-      if (cardinality != 65536) { // all docs are set
+      if (cardinality != BLOCK_SIZE) { // all docs are set
         if (denseRankPower != -1) {
           final byte[] rank = createRank(buffer, denseRankPower);
           out.writeBytes(rank, rank.length);
@@ -418,6 +418,7 @@ public static RandomAccessInput createJumpTable(
 
   // SPARSE variables
   boolean exists;
+  int nextExistDocInBlock = -1;
 
   // DENSE variables
   long word;
@@ -495,7 +496,8 @@ private void readBlockHeader() throws IOException {
     if (numValues <= MAX_ARRAY_LENGTH) {
       method = Method.SPARSE;
       blockEnd = slice.getFilePointer() + (numValues << 1);
-    } else if (numValues == 65536) {
+      nextExistDocInBlock = -1;
+    } else if (numValues == BLOCK_SIZE) {
       method = Method.ALL;
       blockEnd = slice.getFilePointer();
       gap = block - index - 1;
@@ -550,6 +552,7 @@ boolean advanceWithinBlock(IndexedDISI disi, int target) throws IOException {
           if (doc >= targetInBlock) {
             disi.doc = disi.block | doc;
             disi.exists = true;
+            disi.nextExistDocInBlock = doc;
             return true;
           }
         }
@@ -560,13 +563,18 @@ boolean advanceWithinBlock(IndexedDISI disi, int target) throws IOException {
       boolean advanceExactWithinBlock(IndexedDISI disi, int target) throws IOException {
         final int targetInBlock = target & 0xFFFF;
         // TODO: binary search
+        if (disi.nextExistDocInBlock > targetInBlock) {
+          assert !disi.exists;
+          return false;
+        }
         if (target == disi.doc) {
           return disi.exists;
         }
         for (; disi.index < disi.nextBlockIndex; ) {
           int doc = Short.toUnsignedInt(disi.slice.readShort());
           disi.index++;
           if (doc >= targetInBlock) {
+            disi.nextExistDocInBlock = doc;
             if (doc != targetInBlock) {
               disi.index--;
               disi.slice.seek(disi.slice.getFilePointer() - Short.BYTES);