maropu commented on a change in pull request #29085:
URL: https://github.com/apache/spark/pull/29085#discussion_r456440148
##########
File path:
sql/hive/src/main/scala/org/apache/spark/sql/hive/execution/HiveScriptTransformationExec.scala
##########
@@ -54,45 +53,112 @@ case class HiveScriptTransformationExec(
script: String,
output: Seq[Attribute],
child: SparkPlan,
- ioschema: HiveScriptIOSchema)
- extends BaseScriptTransformationExec {
+ ioschema: ScriptTransformationIOSchema)
+ extends BaseScriptTransformationExec with HiveInspectors {
+
+ def initInputSerDe(input: Seq[Expression]): Option[(AbstractSerDe,
StructObjectInspector)] = {
+ ioschema.inputSerdeClass.map { serdeClass =>
+ val (columns, columnTypes) = parseAttrs(input)
+ val serde = initSerDe(serdeClass, columns, columnTypes,
ioschema.inputSerdeProps)
+ val fieldObjectInspectors = columnTypes.map(toInspector)
+ val objectInspector = ObjectInspectorFactory
+ .getStandardStructObjectInspector(columns.asJava,
fieldObjectInspectors.asJava)
+ (serde, objectInspector)
+ }
+ }
+
+ def initOutputSerDe(output: Seq[Attribute]): Option[(AbstractSerDe,
StructObjectInspector)] = {
+ ioschema.outputSerdeClass.map { serdeClass =>
+ val (columns, columnTypes) = parseAttrs(output)
+ val serde = initSerDe(serdeClass, columns, columnTypes,
ioschema.outputSerdeProps)
+ val structObjectInspector =
serde.getObjectInspector().asInstanceOf[StructObjectInspector]
+ (serde, structObjectInspector)
+ }
+ }
+
+ private def parseAttrs(attrs: Seq[Expression]): (Seq[String], Seq[DataType])
= {
+ val columns = attrs.zipWithIndex.map(e => s"${e._1.prettyName}_${e._2}")
+ val columnTypes = attrs.map(_.dataType)
+ (columns, columnTypes)
+ }
+
+ private def initSerDe(
+ serdeClassName: String,
+ columns: Seq[String],
+ columnTypes: Seq[DataType],
+ serdeProps: Seq[(String, String)]): AbstractSerDe = {
+
+ val serde =
Utils.classForName[AbstractSerDe](serdeClassName).getConstructor().
+ newInstance()
+
+ val columnTypesNames =
columnTypes.map(_.toTypeInfo.getTypeName()).mkString(",")
+
+ var propsMap = serdeProps.toMap + (serdeConstants.LIST_COLUMNS ->
columns.mkString(","))
+ propsMap = propsMap + (serdeConstants.LIST_COLUMN_TYPES ->
columnTypesNames)
+
+ val properties = new Properties()
+ // Can not use properties.putAll(propsMap.asJava) in scala-2.12
+ // See https://github.com/scala/bug/issues/10418
+ propsMap.foreach { case (k, v) => properties.put(k, v) }
+ serde.initialize(null, properties)
+
+ serde
+ }
+
+ def recordReader(
Review comment:
private
##########
File path:
sql/hive/src/main/scala/org/apache/spark/sql/hive/execution/HiveScriptTransformationExec.scala
##########
@@ -54,45 +53,112 @@ case class HiveScriptTransformationExec(
script: String,
output: Seq[Attribute],
child: SparkPlan,
- ioschema: HiveScriptIOSchema)
- extends BaseScriptTransformationExec {
+ ioschema: ScriptTransformationIOSchema)
+ extends BaseScriptTransformationExec with HiveInspectors {
+
+ def initInputSerDe(input: Seq[Expression]): Option[(AbstractSerDe,
StructObjectInspector)] = {
+ ioschema.inputSerdeClass.map { serdeClass =>
+ val (columns, columnTypes) = parseAttrs(input)
+ val serde = initSerDe(serdeClass, columns, columnTypes,
ioschema.inputSerdeProps)
+ val fieldObjectInspectors = columnTypes.map(toInspector)
+ val objectInspector = ObjectInspectorFactory
+ .getStandardStructObjectInspector(columns.asJava,
fieldObjectInspectors.asJava)
+ (serde, objectInspector)
+ }
+ }
+
+ def initOutputSerDe(output: Seq[Attribute]): Option[(AbstractSerDe,
StructObjectInspector)] = {
+ ioschema.outputSerdeClass.map { serdeClass =>
+ val (columns, columnTypes) = parseAttrs(output)
+ val serde = initSerDe(serdeClass, columns, columnTypes,
ioschema.outputSerdeProps)
+ val structObjectInspector =
serde.getObjectInspector().asInstanceOf[StructObjectInspector]
+ (serde, structObjectInspector)
+ }
+ }
+
+ private def parseAttrs(attrs: Seq[Expression]): (Seq[String], Seq[DataType])
= {
+ val columns = attrs.zipWithIndex.map(e => s"${e._1.prettyName}_${e._2}")
+ val columnTypes = attrs.map(_.dataType)
+ (columns, columnTypes)
+ }
+
+ private def initSerDe(
+ serdeClassName: String,
+ columns: Seq[String],
+ columnTypes: Seq[DataType],
+ serdeProps: Seq[(String, String)]): AbstractSerDe = {
+
+ val serde =
Utils.classForName[AbstractSerDe](serdeClassName).getConstructor().
+ newInstance()
+
+ val columnTypesNames =
columnTypes.map(_.toTypeInfo.getTypeName()).mkString(",")
+
+ var propsMap = serdeProps.toMap + (serdeConstants.LIST_COLUMNS ->
columns.mkString(","))
+ propsMap = propsMap + (serdeConstants.LIST_COLUMN_TYPES ->
columnTypesNames)
+
+ val properties = new Properties()
+ // Can not use properties.putAll(propsMap.asJava) in scala-2.12
+ // See https://github.com/scala/bug/issues/10418
+ propsMap.foreach { case (k, v) => properties.put(k, v) }
+ serde.initialize(null, properties)
+
+ serde
+ }
+
+ def recordReader(
+ inputStream: InputStream,
+ conf: Configuration): Option[RecordReader] = {
+ ioschema.recordReaderClass.map { klass =>
+ val instance = Utils.classForName[RecordReader](klass).getConstructor().
+ newInstance()
+ val props = new Properties()
+ // Can not use props.putAll(outputSerdeProps.toMap.asJava) in scala-2.12
+ // See https://github.com/scala/bug/issues/10418
+ ioschema.outputSerdeProps.toMap.foreach { case (k, v) => props.put(k, v)
}
+ instance.initialize(inputStream, conf, props)
+ instance
+ }
+ }
+
+ def recordWriter(outputStream: OutputStream, conf: Configuration):
Option[RecordWriter] = {
Review comment:
private
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
