[SPARK-39259][SQL][TEST][FOLLOWUP] Fix Scala 2.13 ClassCastException in ComputeCurrentTimeSuite

### What changes were proposed in this pull request?

Unfortunately, #36654 causes seven Scala 2.13 test failures in master/3.3 and Apache Spark 3.3 RC4.
This PR aims to fix Scala 2.13 ClassCastException in the test code.

### Why are the changes needed?

```
$ dev/change-scala-version.sh 2.13
$ build/sbt "catalyst/testOnly *.ComputeCurrentTimeSuite" -Pscala-2.13
...
[info] ComputeCurrentTimeSuite:
[info] - analyzer should replace current_timestamp with literals *** FAILED *** (1 second, 189 milliseconds)
[info]   java.lang.ClassCastException: scala.collection.mutable.ArrayBuffer cannot be cast to scala.collection.immutable.Seq
[info]   at org.apache.spark.sql.catalyst.optimizer.ComputeCurrentTimeSuite.literals(ComputeCurrentTimeSuite.scala:146)
[info]   at org.apache.spark.sql.catalyst.optimizer.ComputeCurrentTimeSuite.$anonfun$new$1(ComputeCurrentTimeSuite.scala:47)
...
[info] *** 7 TESTS FAILED ***
[error] Failed tests:
[error] 	org.apache.spark.sql.catalyst.optimizer.ComputeCurrentTimeSuite
[error] (catalyst / Test / testOnly) sbt.TestsFailedException: Tests unsuccessful
[error] Total time: 189 s (03:09), completed Jun 3, 2022 10:29:39 AM
```

### Does this PR introduce _any_ user-facing change?

No.

### How was this patch tested?

Pass the CIs and manually tests with Scala 2.13.

```
$ dev/change-scala-version.sh 2.13
$ build/sbt "catalyst/testOnly *.ComputeCurrentTimeSuite" -Pscala-2.13
...
[info] ComputeCurrentTimeSuite:
[info] - analyzer should replace current_timestamp with literals (545 milliseconds)
[info] - analyzer should replace current_date with literals (11 milliseconds)
[info] - SPARK-33469: Add current_timezone function (3 milliseconds)
[info] - analyzer should replace localtimestamp with literals (4 milliseconds)
[info] - analyzer should use equal timestamps across subqueries (182 milliseconds)
[info] - analyzer should use consistent timestamps for different timezones (13 milliseconds)
[info] - analyzer should use consistent timestamps for different timestamp functions (2 milliseconds)
[info] Run completed in 1 second, 579 milliseconds.
[info] Total number of tests run: 7
[info] Suites: completed 1, aborted 0
[info] Tests: succeeded 7, failed 0, canceled 0, ignored 0, pending 0
[info] All tests passed.
[success] Total time: 12 s, completed Jun 3, 2022, 10:54:03 AM
```

Closes #36762 from dongjoon-hyun/SPARK-39259.

Authored-by: Dongjoon Hyun <dongjoon@apache.org>
Signed-off-by: Dongjoon Hyun <dongjoon@apache.org>
(cherry picked from commit d79aa36)
Signed-off-by: Dongjoon Hyun <dongjoon@apache.org>

Author: dongjoon-hyun

sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/ComputeCurrentTimeSuite.scala

@@ -135,14 +135,14 @@ class ComputeCurrentTimeSuite extends PlanTest {
     assert(offsetsFromQuarterHour.size == 1)
   }
 
-  private def literals[T](plan: LogicalPlan): Seq[T] = {
+  private def literals[T](plan: LogicalPlan): scala.collection.mutable.ArrayBuffer[T] = {
     val literals = new scala.collection.mutable.ArrayBuffer[T]
     plan.transformWithSubqueries { case subQuery =>
       subQuery.transformAllExpressions { case expression: Literal =>
         literals += expression.value.asInstanceOf[T]
         expression
       }
     }
-    literals.asInstanceOf[Seq[T]]
+    literals
   }
 }