Github user srowen commented on a diff in the pull request:

    https://github.com/apache/spark/pull/4927#discussion_r25939702
  
    --- Diff: 
core/src/main/scala/org/apache/spark/scheduler/ReplayListenerBus.scala ---
    @@ -40,15 +41,31 @@ private[spark] class ReplayListenerBus extends 
SparkListenerBus with Logging {
        *
        * @param logData Stream containing event log data.
        * @param sourceName Filename (or other source identifier) from whence 
@logData is being read
    +   * @param sourceTruncated Indicate whether log file might be truncated 
(some abnormal situation 
    +   *        encountered, log file not finish writing) or not
        */
    -  def replay(logData: InputStream, sourceName: String): Unit = {
    +  def replay(
    +      logData: InputStream,
    +      sourceName: String,
    +      sourceTruncated: Boolean = false): Unit = {
         var currentLine: String = null
         var lineNumber: Int = 1
         try {
           val lines = Source.fromInputStream(logData).getLines()
    -      lines.foreach { line =>
    -        currentLine = line
    -        postToAll(JsonProtocol.sparkEventFromJson(parse(line)))
    +      while (lines.hasNext) {
    +        currentLine = lines.next()
    +        try {
    +          postToAll(JsonProtocol.sparkEventFromJson(parse(currentLine)))
    +        } catch {
    +          case jpe: JsonParseException =>
    +            // We can only ignore exception from last line of the file 
that might be truncated
    +            if (!sourceTruncated || lines.hasNext) {
    +              throw jpe
    +            } else {
    +              logWarning(s"Get json parse exception from log file 
$sourceName" + 
    --- End diff --
    
    Nit: suggest fixing up the wording a little bit:
    
    ```
    logWarning(s"Got JsonParseException from log file $sourceName" + 
    s" at line $lineNumber; the file might not have finished writing cleanly.")


---
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 infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---

---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org

Reply via email to