Github user davies commented on a diff in the pull request:
https://github.com/apache/spark/pull/1063#discussion_r15712228
--- Diff:
sql/core/src/main/scala/org/apache/spark/sql/execution/pythonUdfs.scala ---
@@ -0,0 +1,182 @@
+/*
+* Licensed to the Apache Software Foundation (ASF) under one or more
+* contributor license agreements. See the NOTICE file distributed with
+* this work for additional information regarding copyright ownership.
+* The ASF licenses this file to You under the Apache License, Version 2.0
+* (the "License"); you may not use this file except in compliance with
+* the License. You may obtain a copy of the License at
+*
+* http://www.apache.org/licenses/LICENSE-2.0
+*
+* Unless required by applicable law or agreed to in writing, software
+* distributed under the License is distributed on an "AS IS" BASIS,
+* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+* See the License for the specific language governing permissions and
+* limitations under the License.
+*/
+
+package org.apache.spark.sql.execution
+
+import java.util.{List => JList, Map => JMap}
+
+import net.razorvine.pickle.{Pickler, Unpickler}
+import org.apache.spark.annotation.DeveloperApi
+import org.apache.spark.api.python.PythonRDD
+import org.apache.spark.broadcast.Broadcast
+import org.apache.spark.rdd.RDD
+import org.apache.spark.sql.catalyst.expressions._
+import org.apache.spark.sql.catalyst.plans.logical
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.types._
+import org.apache.spark.{Accumulator, Logging => SparkLogging}
+
+import scala.collection.JavaConversions._
+
+/**
+ * A serialized version of a Python lambda function. Suitable for use in
a [[PythonRDD]].
+ */
+private[spark] case class PythonUDF(
+ name: String,
+ command: Array[Byte],
+ envVars: JMap[String, String],
+ pythonIncludes: JList[String],
+ pythonExec: String,
+ accumulator: Accumulator[JList[Array[Byte]]],
+ dataType: DataType,
+ children: Seq[Expression]) extends Expression with SparkLogging {
+
+ override def toString = s"PythonUDF#$name(${children.mkString(",")})"
+
+ def nullable: Boolean = true
+ def references: Set[Attribute] = children.flatMap(_.references).toSet
+
+ override def eval(input: Row) = sys.error("PythonUDFs can not be
directly evaluated.")
+}
+
+/**
+ * Extracts PythonUDFs from operators, rewriting the query plan so that
the UDF can be evaluated
+ * alone in a batch.
+ *
+ * This has the limitation that the input to the Python UDF is not allowed
include attributes from
+ * multiple child operators.
+ */
+private[spark] object ExtractPythonUdfs extends Rule[LogicalPlan] {
+ def apply(plan: LogicalPlan) = plan transform {
+ // Skip EvaluatePython nodes.
+ case p: EvaluatePython => p
+
+ case l: LogicalPlan =>
+ // Extract any PythonUDFs from the current operator.
+ val udfs = l.expressions.flatMap(_.collect { case udf: PythonUDF =>
udf})
+ if (udfs.isEmpty) {
+ // If there aren't any, we are done.
+ l
+ } else {
+ // Pick the UDF we are going to evaluate (TODO: Support evaluating
multiple UDFs at a time)
+ // If there is more than one, we will add another evaluation
operator in a subsequent pass.
+ val udf = udfs.head
+
+ var evaluation: EvaluatePython = null
+
+ // Rewrite the child that has the input required for the UDF
+ val newChildren = l.children.map { child =>
+ if (udf.references.subsetOf(child.outputSet)) {
+ evaluation = EvaluatePython(udf, child)
+ evaluation
+ } else if (udf.references.intersect(child.outputSet).nonEmpty) {
+ sys.error(s"Invalid PythonUDF $udf, requires attributes from
more than one child.")
+ } else {
+ child
+ }
+ }
+
+ assert(evaluation != null, "Unable to evaluate PythonUDF. Missing
input attributes.")
+
+ // Trim away the new UDF value if it was only used for filtering
or something.
+ logical.Project(
+ l.output,
+ l.transformExpressions {
+ case p: PythonUDF if p.id == udf.id =>
evaluation.resultAttribute
+ }.withNewChildren(newChildren))
+ }
+ }
+}
+
+/**
+ * :: DeveloperApi ::
+ * Evaluates a [[PythonUDF]], appending the result to the end of the input
tuple.
+ */
+@DeveloperApi
+case class EvaluatePython(udf: PythonUDF, child: LogicalPlan) extends
logical.UnaryNode {
+ val resultAttribute = AttributeReference("pythonUDF", udf.dataType,
nullable=true)()
+
+ def references = Set.empty
+ def output = child.output :+ resultAttribute
+}
+
+/**
+ * :: DeveloperApi ::
+ * Uses PythonRDD to evaluate a [[PythonUDF]], one partition of tuples at
a time. The input
+ * data is cached and zipped with the result of the udf evaluation.
+ */
+@DeveloperApi
+case class BatchPythonEvaluation(udf: PythonUDF, output: Seq[Attribute],
child: SparkPlan)
+ extends SparkPlan {
+ def children = child :: Nil
+
+ def execute() = {
+ // TODO: Clean up after ourselves?
+ val childResults = child.execute().map(_.copy()).cache()
+
+ val parent = childResults.mapPartitions { iter =>
+ val pickle = new Pickler
+ val currentRow = newMutableProjection(udf.children, child.output)()
+ iter.map { inputRow =>
+ val toBePickled = currentRow(inputRow)
+ log.debug(s"toBePickled: $toBePickled")
+ if(children.length == 1) {
+ pickle.dumps(toBePickled.toArray)
+ } else {
+ pickle.dumps(Array(toBePickled.toArray))
+ }
--- End diff --
toBePickled should always be an array.
iter should be grouped in batch before pickling
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
