use Trial for reads

spotify · Sep 19, 2024 · a0b4fe2 · a0b4fe2
1 parent 89695a7
commit a0b4fe2
Show file tree

Hide file tree

Showing 2 changed files with 6 additions and 9 deletions.
diff --git a/jmh/src/test/scala/magnolify/jmh/MagnolifyBench.scala b/jmh/src/test/scala/magnolify/jmh/MagnolifyBench.scala
@@ -211,7 +211,7 @@ object ParquetStates {
 
     var reader: RecordReader[T] = null
 
-    @Setup(Level.Iteration)
+    @Setup(Level.Trial)
     def setup(): Unit = {
       // Write page
       val columnIO = new ColumnIOFactory(true).getColumnIO(schema)
@@ -254,11 +254,7 @@ object ParquetStates {
           new ColumnWriteStoreV1(
             schema,
             new ParquetInMemoryPageStore(Long.MaxValue, writeOnly = true),
-            ParquetProperties.builder
-              .withPageSize(10 * 1024 * 1024)
-              .withPageRowCountLimit(Int.MaxValue)
-              .withDictionaryEncoding(false)
-              .build
+            ParquetProperties.builder.withDictionaryEncoding(false).build
           )
         )
       writeSupport.init(new PlainParquetConfiguration())

diff --git a/jmh/src/test/scala/magnolify/jmh/ParquetInMemoryPageStore.scala b/jmh/src/test/scala/magnolify/jmh/ParquetInMemoryPageStore.scala
@@ -103,6 +103,7 @@ class ParquetInMemoryWriter(writeOnly: Boolean) extends PageWriter {
     dlEncoding: Encoding,
     valuesEncoding: Encoding
   ): Unit = {
+    // If we don't need to read the values later, don't waste mem storing them
     if (!writeOnly) {
       pages.addOne(
         new DataPageV1(
@@ -115,10 +116,10 @@ class ParquetInMemoryWriter(writeOnly: Boolean) extends PageWriter {
           valuesEncoding
         )
       )
-      memSize += bytesInput.size()
-      numRows += rowCount
-      numValues += valueCount
     }
+    memSize += bytesInput.size()
+    numRows += rowCount
+    numValues += valueCount
   }
 
   override def writePageV2(