[SPARK-45026][CONNECT][FOLLOW-UP] Code cleanup

### What changes were proposed in this pull request?
move 3 variables to `isCommand` branch

### Why are the changes needed?
they are not used in other branches

### Does this PR introduce _any_ user-facing change?
No

### How was this patch tested?
CI

### Was this patch authored or co-authored using generative AI tooling?
NO

Closes #42765 from zhengruifeng/SPARK-45026-followup.

Authored-by: Ruifeng Zheng <ruifengz@apache.org>
Signed-off-by: Ruifeng Zheng <ruifengz@apache.org>

connector/connect/server/src/main/scala/org/apache/spark/sql/connect/planner/SparkConnectPlanner.scala

@@ -2464,15 +2464,15 @@ class SparkConnectPlanner(val sessionHolder: SessionHolder) extends Logging {
       case _ => Seq.empty
     }
 
-    // Convert the results to Arrow.
-    val schema = df.schema
-    val maxBatchSize = (SparkEnv.get.conf.get(CONNECT_GRPC_ARROW_MAX_BATCH_SIZE) * 0.7).toLong
-    val timeZoneId = session.sessionState.conf.sessionLocalTimeZone
-
     // To avoid explicit handling of the result on the client, we build the expected input
     // of the relation on the server. The client has to simply forward the result.
     val result = SqlCommandResult.newBuilder()
     if (isCommand) {
+      // Convert the results to Arrow.
+      val schema = df.schema
+      val maxBatchSize = (SparkEnv.get.conf.get(CONNECT_GRPC_ARROW_MAX_BATCH_SIZE) * 0.7).toLong
+      val timeZoneId = session.sessionState.conf.sessionLocalTimeZone
+
       // Convert the data.
       val bytes = if (rows.isEmpty) {
         ArrowConverters.createEmptyArrowBatch(