LUCENE-8593: Specialize single value numeric DV updates

The case when all values are the the same on a numeric field update is common for soft_deletes. With the new infrastucture for buffering DV updates we can gain an easy win by specializing the applied updates if all values are the same.
apache · Dec 6, 2018 · b4e1fe4 · b4e1fe4
1 parent 38cfd0e
commit b4e1fe4
Show file tree

Hide file tree

Showing 5 changed files with 98 additions and 5 deletions.
diff --git a/lucene/core/src/java/org/apache/lucene/index/DocValuesFieldUpdates.java b/lucene/core/src/java/org/apache/lucene/index/DocValuesFieldUpdates.java
@@ -425,4 +425,52 @@ final boolean hasValue() {
       return hasValue;
     }
   }
+
+  static abstract class SingleValueDocValuesFieldUpdates extends DocValuesFieldUpdates {
+
+    protected SingleValueDocValuesFieldUpdates(int maxDoc, long delGen, String field, DocValuesType type) {
+      super(maxDoc, delGen, field, type);
+    }
+
+    @Override
+    void add(int doc, long value) {
+      assert longValue() == value;
+      super.add(doc);
+    }
+
+    @Override
+    void add(int doc, BytesRef value) {
+      assert binaryValue().equals(value);
+      super.add(doc);
+    }
+
+    @Override
+    void add(int docId, Iterator iterator) {
+      throw new UnsupportedOperationException();
+    }
+
+    protected abstract BytesRef binaryValue();
+
+    protected abstract long longValue();
+
+    @Override
+    Iterator iterator() {
+      return new DocValuesFieldUpdates.AbstractIterator(size, docs, delGen) {
+        @Override
+        protected void set(long idx) {
+          // nothing to do;
+        }
+
+        @Override
+        long longValue() {
+          return SingleValueDocValuesFieldUpdates.this.longValue();
+        }
+
+        @Override
+        BytesRef binaryValue() {
+          return SingleValueDocValuesFieldUpdates.this.binaryValue();
+        }
+      };
+    }
+  }
 }
diff --git a/lucene/core/src/java/org/apache/lucene/index/FieldUpdatesBuffer.java b/lucene/core/src/java/org/apache/lucene/index/FieldUpdatesBuffer.java
@@ -183,6 +183,18 @@ boolean isNumeric() {
     return isNumeric;
   }
 
+  boolean hasSingleValue() {
+    // we only do this optimization for numerics so far.
+    return isNumeric && numericValues.length == 1;
+  }
+
+  long getNumericValue(int idx) {
+    if (hasValues != null && hasValues.get(idx) == false) {
+      return 0;
+    }
+    return numericValues[getArrayIndex(numericValues.length, idx)];
+  }
+
   /**
    * Struct like class that is used to iterate over all updates in this buffer
    */

diff --git a/lucene/core/src/java/org/apache/lucene/index/FrozenBufferedUpdates.java b/lucene/core/src/java/org/apache/lucene/index/FrozenBufferedUpdates.java
@@ -484,9 +484,10 @@ private static long applyDocValuesUpdates(BufferedUpdatesStream.SegmentState seg
     long updateCount = 0;
 
     // We first write all our updates private, and only in the end publish to the ReadersAndUpdates */
-    Map<String, DocValuesFieldUpdates> holder = new HashMap<>();
+    final List<DocValuesFieldUpdates> resolvedUpdates = new ArrayList<>();
     for (Map.Entry<String, FieldUpdatesBuffer> fieldUpdate : updates.entrySet()) {
       String updateField = fieldUpdate.getKey();
+      DocValuesFieldUpdates dvUpdates = null;
       FieldUpdatesBuffer value = fieldUpdate.getValue();
       boolean isNumeric = value.isNumeric();
       FieldUpdatesBuffer.BufferedUpdateIterator iterator = value.iterator();
@@ -534,14 +535,19 @@ private static long applyDocValuesUpdates(BufferedUpdatesStream.SegmentState seg
         if (termsEnum.seekExact(bufferedUpdate.termValue)) {
           // we don't need term frequencies for this
           postingsEnum = termsEnum.postings(postingsEnum, PostingsEnum.NONE);
-          DocValuesFieldUpdates dvUpdates = holder.get(updateField);
           if (dvUpdates == null) {
             if (isNumeric) {
-              dvUpdates = new NumericDocValuesFieldUpdates(delGen, updateField, segState.reader.maxDoc());
+              if (value.hasSingleValue()) {
+                dvUpdates = new NumericDocValuesFieldUpdates
+                    .SingleValueNumericDocValuesFieldUpdates(delGen, updateField, segState.reader.maxDoc(),
+                    value.getNumericValue(0));
+              } else {
+                dvUpdates = new NumericDocValuesFieldUpdates(delGen, updateField, segState.reader.maxDoc());
+              }
             } else {
               dvUpdates = new BinaryDocValuesFieldUpdates(delGen, updateField, segState.reader.maxDoc());
             }
-            holder.put(updateField, dvUpdates);
+            resolvedUpdates.add(dvUpdates);
           }
           final IntConsumer docIdConsumer;
           final DocValuesFieldUpdates update = dvUpdates;
@@ -582,7 +588,7 @@ private static long applyDocValuesUpdates(BufferedUpdatesStream.SegmentState seg
     }
 
     // now freeze & publish:
-    for (DocValuesFieldUpdates update : holder.values()) {
+    for (DocValuesFieldUpdates update : resolvedUpdates) {
       if (update.any()) {
         update.finish();
         segState.rld.addDVUpdate(update);

diff --git a/lucene/core/src/java/org/apache/lucene/index/NumericDocValuesFieldUpdates.java b/lucene/core/src/java/org/apache/lucene/index/NumericDocValuesFieldUpdates.java
@@ -111,4 +111,24 @@ public long ramBytesUsed() {
         + Long.BYTES
         + RamUsageEstimator.NUM_BYTES_OBJECT_REF;
   }
+
+  static class SingleValueNumericDocValuesFieldUpdates extends SingleValueDocValuesFieldUpdates {
+
+    private final long value;
+
+    SingleValueNumericDocValuesFieldUpdates(long delGen, String field, int maxDoc, long value) {
+      super(maxDoc, delGen, field, DocValuesType.NUMERIC);
+      this.value = value;
+    }
+
+    @Override
+    protected BytesRef binaryValue() {
+      throw new UnsupportedOperationException();
+    }
+
+    @Override
+    protected long longValue() {
+      return value;
+    }
+  }
 }
diff --git a/lucene/core/src/test/org/apache/lucene/index/TestFieldUpdatesBuffer.java b/lucene/core/src/test/org/apache/lucene/index/TestFieldUpdatesBuffer.java
@@ -36,9 +36,13 @@ public void testBascis() throws IOException {
         new DocValuesUpdate.NumericDocValuesUpdate(new Term("id", "1"), "age", 6);
     FieldUpdatesBuffer buffer = new FieldUpdatesBuffer(counter, update, 15);
     buffer.addUpdate(new Term("id", "10"), 6, 15);
+    assertTrue(buffer.hasSingleValue());
     buffer.addUpdate(new Term("id", "8"), 12, 15);
+    assertFalse(buffer.hasSingleValue());
     buffer.addUpdate(new Term("some_other_field", "8"), 13, 17);
+    assertFalse(buffer.hasSingleValue());
     buffer.addUpdate(new Term("id", "8"), 12, 16);
+    assertFalse(buffer.hasSingleValue());
     assertTrue(buffer.isNumeric());
     FieldUpdatesBuffer.BufferedUpdateIterator iterator = buffer.iterator();
     FieldUpdatesBuffer.BufferedUpdate value = iterator.next();
@@ -214,14 +218,17 @@ public void testNumericRandom() throws IOException {
 
     int count = 0;
     while ((value = iterator.next()) != null) {
+      long v = buffer.getNumericValue(count);
       randomUpdate = updates.get(count++);
       assertEquals(randomUpdate.term.bytes.utf8ToString(), value.termValue.utf8ToString());
       assertEquals(randomUpdate.term.field, value.termField);
       assertEquals(randomUpdate.hasValue, value.hasValue);
       if (randomUpdate.hasValue) {
         assertEquals(randomUpdate.getValue(), value.numericValue);
+        assertEquals(v, value.numericValue);
       } else {
         assertEquals(0, value.numericValue);
+        assertEquals(0, v);
       }
       assertEquals(randomUpdate.docIDUpto, value.docUpTo);
     }