[spark] branch master updated: [SPARK-27243][SQL] RuleExecutor.dumpTimeSpent should not throw exception when empty

[gu...@apache.org]

This is an automated email from the ASF dual-hosted git repository.

gurwls223 pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/spark.git


The following commit(s) were added to refs/heads/master by this push:
     new fe317dc  [SPARK-27243][SQL] RuleExecutor.dumpTimeSpent should not throw exception when empty
fe317dc is described below

commit fe317dc74e5fa1509ae9d735485f66724f7292e5
Author: Marco Gaido <ma...@gmail.com>
AuthorDate: Sat Mar 23 09:49:20 2019 +0900

    [SPARK-27243][SQL] RuleExecutor.dumpTimeSpent should not throw exception when empty
    
    ## What changes were proposed in this pull request?
    
    `RuleExecutor.dumpTimeSpent` currently throws an exception when invoked before any rule is run or immediately after `RuleExecutor.reset`. The PR makes it returning an empty summary, which is the expected output instead.
    
    ## How was this patch tested?
    
    added UT
    
    Closes #24180 from mgaido91/SPARK-27243.
    
    Authored-by: Marco Gaido <ma...@gmail.com>
    Signed-off-by: Hyukjin Kwon <gu...@apache.org>
---
 .../apache/spark/sql/catalyst/rules/QueryExecutionMetering.scala    | 6 +++++-
 .../org/apache/spark/sql/catalyst/trees/RuleExecutorSuite.scala     | 6 ++++++
 2 files changed, 11 insertions(+), 1 deletion(-)

diff --git a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/rules/QueryExecutionMetering.scala b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/rules/QueryExecutionMetering.scala
index e4d5fa9..7a86433 100644
--- a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/rules/QueryExecutionMetering.scala
+++ b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/rules/QueryExecutionMetering.scala
@@ -64,7 +64,11 @@ case class QueryExecutionMetering() {
   /** Dump statistics about time spent running specific rules. */
   def dumpTimeSpent(): String = {
     val map = timeMap.asMap().asScala
-    val maxLengthRuleNames = map.keys.map(_.toString.length).max
+    val maxLengthRuleNames = if (map.isEmpty) {
+      0
+    } else {
+      map.keys.map(_.toString.length).max
+    }
 
     val colRuleName = "Rule".padTo(maxLengthRuleNames, " ").mkString
     val colRunTime = "Effective Time / Total Time".padTo(len = 47, " ").mkString
diff --git a/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/trees/RuleExecutorSuite.scala b/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/trees/RuleExecutorSuite.scala
index ab5d722..8dbe198 100644
--- a/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/trees/RuleExecutorSuite.scala
+++ b/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/trees/RuleExecutorSuite.scala
@@ -91,4 +91,10 @@ class RuleExecutorSuite extends SparkFunSuite {
     }.getMessage
     assert(message.contains("the structural integrity of the plan is broken"))
   }
+
+  test("SPARK-27243: dumpTimeSpent when no rule has run") {
+    RuleExecutor.resetMetrics()
+    // This should not throw an exception
+    RuleExecutor.dumpTimeSpent()
+  }
 }


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@spark.apache.org
For additional commands, e-mail: commits-help@spark.apache.org