doc

Author: thunterdb

mllib/src/main/scala/org/apache/spark/ml/stat/Summarizer.scala

@@ -393,6 +393,7 @@ object SummaryBuilderImpl extends Logging {
      * Reads a buffer from a serialized form, using the row object as an assistant.
      */
     def read(bytes: Array[Byte]): Buffer = {
+      // TODO move this row outside to the aggregate
       assert(numFields == 12, numFields)
       val row3 = new UnsafeRow(numFields)
       row3.pointTo(bytes.clone(), bytes.length)

sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/Projection.scala

@@ -101,6 +101,8 @@ case class InterpretedMutableProjection(expressions: Seq[Expression]) extends Mu
 
 /**
  * A projection that returns UnsafeRow.
+ *
+ * CAUTION: the returned projection object should *not* be assumed to be thread-safe.
  */
 abstract class UnsafeProjection extends Projection {
   override def apply(row: InternalRow): UnsafeRow
@@ -110,11 +112,15 @@ object UnsafeProjection {
 
   /**
    * Returns an UnsafeProjection for given StructType.
+   *
+   * CAUTION: the returned projection object is *not* thread-safe.
    */
   def create(schema: StructType): UnsafeProjection = create(schema.fields.map(_.dataType))
 
   /**
    * Returns an UnsafeProjection for given Array of DataTypes.
+   *
+   * CAUTION: the returned projection object is *not* thread-safe.
    */
   def create(fields: Array[DataType]): UnsafeProjection = {
     create(fields.zipWithIndex.map(x => BoundReference(x._2, x._1, true)))