Writing trailing bytes instead

dnhatn · dnhatn · commit 4ee7a6bd90d2 · 2021-10-17T21:35:02.000-04:00
diff --git a/lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90DocValuesProducer.java b/lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90DocValuesProducer.java
@@ -1351,7 +1351,6 @@ public int docValueCount() {
 
         private void set() {
           if (set == false) {
-            assert docID() != NO_MORE_DOCS;
             final int index = disi.index();
             start = addresses.get(index);
             end = addresses.get(index + 1L);
@@ -1403,23 +1402,15 @@ public int docID() {
       @Override
       public int nextDoc() throws IOException {
         int doc = ords.nextDoc();
-        if (doc != NO_MORE_DOCS) {
-          count = ords.docValueCount();
-        } else {
-          count = 0;
-        }
+        count = ords.docValueCount();
         i = 0;
         return doc;
       }
 
       @Override
       public int advance(int target) throws IOException {
         int doc = ords.advance(target);
-        if (doc != NO_MORE_DOCS) {
-          count = ords.docValueCount();
-        } else {
-          count = 0;
-        }
+        count = ords.docValueCount();
         i = 0;
         return doc;
       }
diff --git a/lucene/core/src/java/org/apache/lucene/util/packed/DirectWriter.java b/lucene/core/src/java/org/apache/lucene/util/packed/DirectWriter.java
@@ -149,13 +149,40 @@ public void finish() throws IOException {
     }
     assert !finished;
     flush();
-    // pad for fast io: we actually only need this for certain BPV, but its just 3 bytes...
-    for (int i = 0; i < 3; i++) {
+
+    // padding for fast io and allow over read the last index by 1.
+    final int paddingBytes = numberOfTrailingBytes(bitsPerValue);
+    for (int i = 0; i < paddingBytes; i++) {
       output.writeByte((byte) 0);
     }
     finished = true;
   }
 
+  private static int numberOfTrailingBytes(int bitsPerValue) {
+    switch (bitsPerValue) {
+      case 1:
+      case 2:
+      case 4:
+      case 8:
+        return 1;
+      case 12:
+      case 16:
+        return 2;
+      case 20:
+      case 24:
+      case 28:
+      case 32:
+        return 5;
+      case 40:
+      case 48:
+      case 56:
+      case 64:
+        return 11;
+      default:
+        throw new IllegalArgumentException("unsupported bitsPerValue: " + bitsPerValue);
+    }
+  }
+
   /** Returns an instance suitable for encoding {@code numValues} using {@code bitsPerValue} */
   public static DirectWriter getInstance(DataOutput output, long numValues, int bitsPerValue) {
     if (Arrays.binarySearch(SUPPORTED_BITS_PER_VALUE, bitsPerValue) < 0) {
diff --git a/lucene/core/src/test/org/apache/lucene/util/packed/TestDirectPacked.java b/lucene/core/src/test/org/apache/lucene/util/packed/TestDirectPacked.java
@@ -116,6 +116,8 @@ private void doTestBpv(Directory directory, int bpv, long offset) throws Excepti
       for (int j = 0; j < original.length; j++) {
         assertEquals("bpv=" + bpv, original[j], reader.get(j));
       }
+      // Make sure that we can overread the last index by one.
+      assertEquals("bpv=" + bpv, 0, reader.get(original.length));
       input.close();
     }
   }

Original file line number	Diff line number	Diff line change
`@@ -116,6 +116,8 @@ private void doTestBpv(Directory directory, int bpv, long offset) throws Excepti`
`116`	`116`	`for (int j = 0; j < original.length; j++) {`
`117`	`117`	`assertEquals("bpv=" + bpv, original[j], reader.get(j));`
`118`	`118`	`}`
	`119`	`+ // Make sure that we can overread the last index by one.`
	`120`	`+ assertEquals("bpv=" + bpv, 0, reader.get(original.length));`
`119`	`121`	`input.close();`
`120`	`122`	`}`
`121`	`123`	`}`