Github user pwendell commented on a diff in the pull request:
https://github.com/apache/incubator-spark/pull/625#discussion_r9956570
--- Diff: core/src/main/scala/org/apache/spark/deploy/ClientArguments.scala
---
@@ -115,3 +110,7 @@ private[spark] class ClientArguments(args:
Array[String]) {
System.exit(exitCode)
}
}
+
+object ClientArguments {
+ def isValidJarUrl(s: String) = s.matches("(.+):(.+)jar")
--- End diff --
Hey @hsaputra - I'd think we should recommend "//" to users since it's the
widely used way. Unfortunately URL naming is totally messed up and inconsistent
not only in Hadoop but in the broader internet ecosystem. The recommended
philosphy in the IETF is to be "liberal in what you accept, and conservative in
what you send". I think we should be liberal in accepting weird URL's, but we
should not advise users to make quirky URL's.
http://en.wikipedia.org/wiki/Robustness_principle
That said, this PR is mainly because java.net.URL (incorrectly) considers
`hdfs` an invalid scheme. So we can't rely on java.net.URL to validate this.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. To do so, please top-post your response.
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
infrastruct...@apache.org or file a JIRA ticket with INFRA.
---
