Profiling tool - fix reporting contains dataset when sql time 0

tools/src/main/scala/com/nvidia/spark/rapids/tool/profiling/ClassWarehouse.scala

@@ -57,6 +57,7 @@ case class SQLExecutionCase(
     duration: Option[Long],
     durationStr: String,
     sqlQualDuration: Option[Long],
+    hasDataset: Boolean,
     problematic: String = "")
 
 case class SQLPlanMetricsCase(

tools/src/main/scala/org/apache/spark/sql/rapids/tool/profiling/ApplicationInfo.scala

@@ -347,10 +347,10 @@ class ApplicationInfo(
           case Some(i) => UIUtils.formatDuration(i)
           case None => ""
         }
-        val sqlQDuration = if (datasetSQL.exists(_.sqlID == res.sqlID)) {
-          Some(0L)
+        val (containsDataset, sqlQDuration) = if (datasetSQL.exists(_.sqlID == res.sqlID)) {
+          (true, Some(0L))
         } else {
-          durationResult
+          (false, durationResult)
         }
         val potProbs = problematicSQL.filter { p =>
           p.sqlID == res.sqlID && p.reason.nonEmpty
@@ -364,6 +364,7 @@ class ApplicationInfo(
           duration = durationResult,
           durationStr = durationString,
           sqlQualDuration = sqlQDuration,
+          hasDataset = containsDataset,
           problematic = finalPotProbs
         )
         sqlStartNew += sqlExecutionNew
@@ -851,7 +852,7 @@ class ApplicationInfo(
        |'$appId' as `App ID`,
        |sq.sqlID,
        |sq.duration as `SQL Duration`,
-       |case when sq.sqlQualDuration > 0 then false else true end as `Contains Dataset Op`,
+       |sq.hasDataset as `Contains Dataset Op`,
        |app.duration as `App Duration`,
        |problematic as `Potential Problems`,
        |round(executorCPUTime/executorRunTime*100,2) as `Executor CPU Time Percent`

tools/src/main/scala/org/apache/spark/sql/rapids/tool/profiling/EventsProcessor.scala

@@ -343,6 +343,7 @@ class EventsProcessor(forQualification: Boolean = false) extends Logging {
       None,
       "",
       None,
+      false,
       ""
     )
     app.sqlStart += sqlExecution

tools/src/test/scala/com/nvidia/spark/rapids/tool/profiling/AnalysisSuite.scala

@@ -149,4 +149,22 @@ class AnalysisSuite extends FunSuite {
     val actualDf = analysis.shuffleSkewCheckSingleApp(apps.head)
     assert(actualDf.count() == 0)
   }
+
+  test("test contains dataset false") {
+    val qualLogDir = ToolTestUtils.getTestResourcePath("spark-events-qualification")
+    val logs = Array(s"$qualLogDir/nds_q86_test")
+
+    val apps = ToolTestUtils.processProfileApps(logs, sparkSession)
+    val analysis = new Analysis(apps, None)
+    val sqlAggMetricsDF = analysis.sqlMetricsAggregation()
+    sqlAggMetricsDF.createOrReplaceTempView("sqlAggMetricsDF")
+    val actualDf = analysis.sqlMetricsAggregationDurationAndCpuTime()
+
+    val rows = actualDf.collect()
+    assert(rows.length === 25)
+    def fieldIndex(name: String) = actualDf.schema.fieldIndex(name)
+    rows.foreach { row =>
+      assert(row.getBoolean(fieldIndex("Contains Dataset Op")) == false)
+    }
+  }
 }