Github user srowen commented on a diff in the pull request:
https://github.com/apache/spark/pull/22723#discussion_r229721018
--- Diff:
core/src/main/scala/org/apache/spark/input/WholeTextFileInputFormat.scala ---
@@ -48,11 +50,11 @@ private[spark] class WholeTextFileInputFormat
* Allow minPartitions set by end-user in order to keep compatibility
with old Hadoop API,
* which is set through setMaxSplitSize
*/
- def setMinPartitions(context: JobContext, minPartitions: Int) {
+ def setMinPartitions(sc: SparkContext, context: JobContext,
minPartitions: Int) {
val files = listStatus(context).asScala
val totalLen = files.map(file => if (file.isDirectory) 0L else
file.getLen).sum
- val maxSplitSize = Math.ceil(totalLen * 1.0 /
- (if (minPartitions == 0) 1 else minPartitions)).toLong
+ val minPartNum = Math.max(sc.defaultParallelism, minPartitions)
+ val maxSplitSize = Math.ceil(totalLen * 1.0 / minPartNum).toLong
--- End diff --
Yes, this no longer matches the title or JIRA. I am also not clear on the
argument why this is better?
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
