[jira] [Comment Edited] (SPARK-2331) SparkContext.emptyRDD has wrong return type

["Sean Owen (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/SPARK-2331?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14048041#comment-14048041 ] 

Sean Owen edited comment on SPARK-2331 at 6/30/14 7:48 PM:
-----------------------------------------------------------

My 2 cents. 

{code}
val rdds = Seq("a","b","c").foldLeft(sc.emptyRDD){ (rdd,path) => rdd.union(sc.textFile(path)) }
{code}

yields

{code}
<console>:12: error: type mismatch;
 found   : org.apache.spark.rdd.RDD[String]
 required: org.apache.spark.rdd.RDD[Nothing]
Note: String >: Nothing, but class RDD is invariant in type T.
You may wish to define T as -T instead. (SLS 4.5)
       val rdds = Seq("a","b","c").foldLeft(sc.emptyRDD){ (rdd,path) => rdd.union(sc.textFile(path)) }
{code}

but even

{code}
val rdds = Seq("a","b","c").foldLeft(sc.emptyRDD[String]){ (rdd,path) => rdd.union(sc.textFile(path)) }
{code}

yields

{code}
<console>:12: error: type mismatch;
 found   : org.apache.spark.rdd.RDD[String]
 required: org.apache.spark.rdd.EmptyRDD[String]
       val rdds = Seq("a","b","c").foldLeft(sc.emptyRDD[String]){ (rdd,path) => rdd.union(sc.textFile(path)) }
{code}

So I think this is really about RDDs being invariant, rather than the return type here, and that seems to be how it's going to be: https://issues.apache.org/jira/browse/SPARK-1296

I think there's an argument for hiding EmptyRDD although that would be a little API change at this point.




was (Author: srowen):
My 2 cents. You mean the type is EmptyRDD[Nothing] right? 

{code}
val rdds = Seq("a","b","c").foldLeft(sc.emptyRDD){ (rdd,path) => rdd.union(sc.textFile(path)) }
{code}

yields

{code}
<console>:12: error: type mismatch;
 found   : org.apache.spark.rdd.RDD[String]
 required: org.apache.spark.rdd.RDD[Nothing]
Note: String >: Nothing, but class RDD is invariant in type T.
You may wish to define T as -T instead. (SLS 4.5)
       val rdds = Seq("a","b","c").foldLeft(sc.emptyRDD){ (rdd,path) => rdd.union(sc.textFile(path)) }
{code}

but even

{code}
val rdds = Seq("a","b","c").foldLeft(sc.emptyRDD[String]){ (rdd,path) => rdd.union(sc.textFile(path)) }
{code}

yields

{code}
<console>:12: error: type mismatch;
 found   : org.apache.spark.rdd.RDD[String]
 required: org.apache.spark.rdd.EmptyRDD[String]
       val rdds = Seq("a","b","c").foldLeft(sc.emptyRDD[String]){ (rdd,path) => rdd.union(sc.textFile(path)) }
{code}

So I think this is really about RDDs being invariant, rather than the return type here, and that seems to be how it's going to be: https://issues.apache.org/jira/browse/SPARK-1296

I think there's an argument for hiding EmptyRDD although that would be a little API change at this point.



> SparkContext.emptyRDD has wrong return type
> -------------------------------------------
>
>                 Key: SPARK-2331
>                 URL: https://issues.apache.org/jira/browse/SPARK-2331
>             Project: Spark
>          Issue Type: Bug
>    Affects Versions: 1.0.0
>            Reporter: Ian Hummel
>
> The return type for SparkContext.emptyRDD is EmptyRDD[T].
> It should be RDD[T].  That means you have to add extra type annotations on code like the below (which creates a union of RDDs over some subset of paths in a folder)
> val rdds = Seq("a", "b", "c").foldLeft[RDD[String]](sc.emptyRDD[String]) { (rdd, path) ⇒
>           rdd.union(sc.textFile(path))
>         }



--
This message was sent by Atlassian JIRA
(v6.2#6252)