Fixed bitpacking zero values (#731)

flow-php · Nov 5, 2023 · 3fef49f · 3fef49f
1 parent e6077b3
commit 3fef49f
Show file tree

Hide file tree

Showing 2 changed files with 23 additions and 0 deletions.
diff --git a/src/lib/parquet/src/Flow/Parquet/ParquetFile/Data/RLEBitPackedHybrid.php b/src/lib/parquet/src/Flow/Parquet/ParquetFile/Data/RLEBitPackedHybrid.php
@@ -23,6 +23,12 @@ public function decodeBitPacked(BinaryReader $reader, int $bitWidth, int $varInt
             return;
         }
 
+        if ($bitWidth === 0) {
+            $output = \array_merge($output, \array_fill(0, \min($numGroups * 8, $maxItems), 0));
+
+            return;
+        }
+
         $count = $numGroups * 8;
         $totalByteCount = (int) (($bitWidth * $count) / 8);
 

diff --git a/.../parquet/tests/Flow/Parquet/Tests/Integration/ParquetFile/Data/RLEBitPackedHybridTest.php b/.../parquet/tests/Flow/Parquet/Tests/Integration/ParquetFile/Data/RLEBitPackedHybridTest.php
@@ -63,6 +63,23 @@ public function test_bit_packing_with_two_sequences() : void
         );
     }
 
+    public function test_bit_packing_zeroes() : void
+    {
+        $values = [0, 0, 0, 0];
+
+        $buffer = '';
+        (new RLEBitPackedHybrid())->encodeHybrid(new BinaryBufferWriter($buffer), $values);
+
+        $this->assertSame(
+            $values,
+            (new RLEBitPackedHybrid())->decodeHybrid(
+                new BinaryBufferReader($buffer),
+                BitWidth::fromArray($values),
+                \count($values)
+            )
+        );
+    }
+
     public function test_bit_rle_reading_with_extra_bytes_in_the_buffer() : void
     {
         $values = [3, 3, 3, 3, 3, 3, 3, 3, 3, 3];