fix: Update Spark test expectation for TIMESTAMPTZ values

After rebasing on main, the Spark test expectations were reverted to the
old incorrect values. This commit re-applies the fix for TIMESTAMPTZ
handling in HyperResultSourceTest.

The test now correctly expects Timestamp.from(Instant.parse(...)) for
TIMESTAMPTZ columns, which preserves the correct epoch milliseconds
instead of incorrectly interpreting the value as a local time.

Test: HyperResultSourceTest - supports reading all types

Author: KaviarasuSakthivadivel

spark-datasource-core/src/test/scala/HyperResultSourceTest.scala

@@ -8,6 +8,7 @@ import com.salesforce.datacloud.jdbc.hyper.HyperServerManager.ConfigFile
 
 import java.sql.{Date, SQLException, Timestamp}
 import java.math.BigDecimal
+import java.time.Instant
 import org.apache.spark.sql.types.{
   BinaryType,
   BooleanType,
@@ -214,8 +215,8 @@ class HyperResultSourceTest extends AnyFunSuite with WithSparkSession {
     assert(row.getAs[Date]("date") == Date.valueOf("2024-01-01"))
     assert(row.isNullAt(row.fieldIndex("date_null")))
     assert(
-      row.getAs[Timestamp]("timestamp") == Timestamp.valueOf(
-        "2024-01-01 12:00:00"
+      row.getAs[Timestamp]("timestamp") == Timestamp.from(
+        Instant.parse("2024-01-02T04:00:00Z")
       )
     )
     assert(