Github user srowen commented on a diff in the pull request:
https://github.com/apache/spark/pull/13351#discussion_r64901932
--- Diff: core/src/main/scala/org/apache/spark/util/Utils.scala ---
@@ -2345,17 +2345,19 @@ private[spark] class RedirectThread(
*/
private[spark] class CircularBuffer(sizeInBytes: Int = 10240) extends
java.io.OutputStream {
var pos: Int = 0
+ var isBufferFull = false
var buffer = new Array[Int](sizeInBytes)
def write(i: Int): Unit = {
buffer(pos) = i
pos = (pos + 1) % buffer.length
+ isBufferFull = isBufferFull || (pos == 0)
}
override def toString: String = {
val (end, start) = buffer.splitAt(pos)
val input = new java.io.InputStream {
- val iterator = (start ++ end).iterator
+ val iterator = (if (isBufferFull) start ++ end else end).iterator
--- End diff --
This also becomes fairly inefficient, since it copies the entire buffer to
make the two halves.
Since we're already copying and trying to parse this as a string, rather
than make an input stream that reads byte by byte, how about just converting
one or both halves of the byte array to a string and reading it with
`StringReader`?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. 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 [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
