[SPARK-19004][SQL] Fix `JDBCWriteSuite.testH2Dialect` by removing `getCatalystType`

## What changes were proposed in this pull request?

`JDBCSuite` and `JDBCWriterSuite` have their own `testH2Dialect`s for their testing purposes.

This PR fixes `testH2Dialect` in `JDBCWriterSuite` by removing `getCatalystType` implementation in order to return correct types. Currently, it always returns `Some(StringType)` incorrectly. Note that, for the `testH2Dialect` in `JDBCSuite`, it's intentional because of the test case `Remap types via JdbcDialects`.

## How was this patch tested?

This is a test only update.

Author: Dongjoon Hyun <dongjoon@apache.org>

Closes #16409 from dongjoon-hyun/SPARK-H2-DIALECT.

sql/core/src/test/scala/org/apache/spark/sql/jdbc/JDBCWriteSuite.scala

@@ -44,9 +44,6 @@ class JDBCWriteSuite extends SharedSQLContext with BeforeAndAfter {
 
   val testH2Dialect = new JdbcDialect {
     override def canHandle(url: String) : Boolean = url.startsWith("jdbc:h2")
-    override def getCatalystType(
-        sqlType: Int, typeName: String, size: Int, md: MetadataBuilder): Option[DataType] =
-      Some(StringType)
     override def isCascadingTruncateTable(): Option[Boolean] = Some(false)
   }