Github user ifilonenko commented on a diff in the pull request:
https://github.com/apache/spark/pull/22911#discussion_r233542062
--- Diff:
resource-managers/kubernetes/core/src/main/scala/org/apache/spark/deploy/k8s/SparkPod.scala
---
@@ -18,7 +18,30 @@ package org.apache.spark.deploy.k8s
import io.fabric8.kubernetes.api.model.{Container, ContainerBuilder, Pod,
PodBuilder}
-private[spark] case class SparkPod(pod: Pod, container: Container)
+private[spark] case class SparkPod(pod: Pod, container: Container) {
+
+ /**
+ * Convenience method to apply a series of chained transformations to a
pod.
+ *
+ * Use it like:
+ *
+ * original.modify { case pod =>
+ * // update pod and return new one
+ * }.modify { case pod =>
+ * // more changes that create a new pod
+ * }.modify {
+ * case pod if someCondition => // new pod
+ * }
+ *
+ * This makes it cleaner to apply multiple transformations, avoiding
having to create
+ * a bunch of awkwardly-named local variables. Since the argument is a
partial function,
+ * it can do matching without needing to exhaust all the possibilities.
If the function
+ * is not applied, then the original pod will be kept.
+ */
+ def transform(fn: PartialFunction[SparkPod, SparkPod]): SparkPod =
fn.lift(this).getOrElse(this)
--- End diff --
I would think that this change is out of the scope of this PR, but I do
love the use of a PartialFunction here. Thanks for this!
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
