[SPARK-37654][SQL] Fix NPE in Row.getSeq when field is Null

### What changes were proposed in this pull request?
Fix NPE
```
scala> Row(null).getSeq(0)
java.lang.NullPointerException
  at org.apache.spark.sql.Row.getSeq(Row.scala:319)
  at org.apache.spark.sql.Row.getSeq$(Row.scala:319)
  at org.apache.spark.sql.catalyst.expressions.GenericRow.getSeq(rows.scala:166)
```

### Why are the changes needed?
bug fixing

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

### How was this patch tested?
new UT

Closes #34928 from huaxingao/npe.

Authored-by: Huaxin Gao <[email protected]>
Signed-off-by: Hyukjin Kwon <[email protected]>

Author: huaxingao

sql/catalyst/src/main/scala/org/apache/spark/sql/Row.scala

@@ -317,7 +317,10 @@ trait Row extends Serializable {
    *
    * @throws ClassCastException when data type does not match.
    */
-  def getSeq[T](i: Int): Seq[T] = getAs[scala.collection.Seq[T]](i).toSeq
+  def getSeq[T](i: Int): Seq[T] = {
+    val res = getAs[scala.collection.Seq[T]](i)
+    if (res != null) res.toSeq else null
+  }
 
   /**
    * Returns the value at position i of array type as `java.util.List`.

sql/core/src/test/scala/org/apache/spark/sql/RowSuite.scala

@@ -105,4 +105,10 @@ class RowSuite extends SparkFunSuite with SharedSparkSession {
     val empty = Row()
     assert(empty.toString == "[]")
   }
+
+  test("SPARK-37654: row contains a null at the requested index should return null") {
+    assert(Row(Seq("value")).getSeq(0) === List("value"))
+    assert(Row(Seq()).getSeq(0) === List())
+    assert(Row(null).getSeq(0) === null)
+  }
 }