[SPARK-53125][TEST] RemoteSparkSession prints whole spark-submit command

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

Make RemoteSparkSession print the whole `spark-submit` command in debug mode, which helps the developers to understand the SPARK_HOME, classpath, log output, etc. of the testing connect server process.

### Why are the changes needed?

Improve the debug message.

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

No.

### How was this patch tested?

```
export SPARK_DEBUG_SC_JVM_CLIENT=true
```
```
sbt:spark-connect-client-jvm> testOnly *ClientE2ETestSuite -- -z "throw SparkException with large cause exception"
[info] ClientE2ETestSuite:
Starting the Spark Connect Server...
Using jar: /Users/chengpan/Projects/apache-spark/sql/connect/server/target/scala-2.13/spark-connect-assembly-4.1.0-SNAPSHOT.jar
Using jar: /Users/chengpan/Projects/apache-spark/sql/catalyst/target/scala-2.13/spark-catalyst_2.13-4.1.0-SNAPSHOT-tests.jar
/Users/chengpan/Projects/apache-spark/bin/spark-submit \
    --driver-class-path /Users/chengpan/Projects/apache-spark/sql/connect/server/target/scala-2.13/spark-connect-assembly-4.1.0-SNAPSHOT.jar \
    --class org.apache.spark.sql.connect.SimpleSparkConnectService \
    --jars /Users/chengpan/Projects/apache-spark/sql/catalyst/target/scala-2.13/spark-catalyst_2.13-4.1.0-SNAPSHOT-tests.jar \
    --conf spark.connect.grpc.binding.port=15707 \
    --conf spark.sql.catalog.testcat=org.apache.spark.sql.connector.catalog.InMemoryTableCatalog \
    --conf spark.sql.catalogImplementation=hive \
    --conf spark.connect.execute.reattachable.senderMaxStreamDuration=1s \
    --conf spark.connect.execute.reattachable.senderMaxStreamSize=123 \
    --conf spark.connect.grpc.arrow.maxBatchSize=10485760 \
    --conf spark.ui.enabled=false \
    --conf spark.driver.extraJavaOptions=-Dlog4j.configurationFile=/Users/chengpan/Projects/apache-spark/sql/connect/client/jvm/src/test/resources/log4j2.properties /Users/chengpan/Projects/apache-spark/sql/connect/server/target/scala-2.13/spark-connect-assembly-4.1.0-SNAPSHOT.jar
...
```

### Was this patch authored or co-authored using generative AI tooling?

No.

Closes #51846 from pan3793/SPARK-53125.

Authored-by: Cheng Pan <[email protected]>
Signed-off-by: Dongjoon Hyun <[email protected]>

Author: pan3793

sql/connect/client/jvm/src/test/scala/org/apache/spark/sql/connect/test/RemoteSparkSession.scala

@@ -71,15 +71,22 @@ object SparkConnectServerUtils {
       findJar("sql/catalyst", "spark-catalyst", "spark-catalyst", test = true).getCanonicalPath
 
     val command = Seq.newBuilder[String]
-    command += "bin/spark-submit"
+    command += s"$sparkHome/bin/spark-submit"
     command += "--driver-class-path" += connectJar
     command += "--class" += "org.apache.spark.sql.connect.SimpleSparkConnectService"
     command += "--jars" += catalystTestJar
     command += "--conf" += s"spark.connect.grpc.binding.port=$port"
     command ++= testConfigs
     command ++= debugConfigs
     command += connectJar
-    val builder = new ProcessBuilder(command.result(): _*)
+    val cmds = command.result()
+    debug {
+      cmds.reduce[String] {
+        case (acc, cmd) if cmd startsWith "-" => acc + " \\\n    " + cmd
+        case (acc, cmd) => acc + " " + cmd
+      }
+    }
+    val builder = new ProcessBuilder(cmds: _*)
     builder.directory(new File(sparkHome))
     val environment = builder.environment()
     environment.remove("SPARK_DIST_CLASSPATH")