[ARROW] Update arguments of ArrowUtils#toArrowSchema function - #4923

Why are the changes needed?

to adapt Spark 3.5, the new conf spark.sql.execution.arrow.useLargeVarType was introduced in apache/spark#39572

the signature of function ArrowUtils#toArrowSchema before

   def toArrowSchema(
       schema: StructType,
       timeZoneId: String,
       errorOnDuplicatedFieldNames: Boolean): Schema

after

  def toArrowSchema(
      schema: StructType,
      timeZoneId: String,
      errorOnDuplicatedFieldNames: Boolean,
      largeVarTypes: Boolean = false): Schema

How was this patch tested?

• Add some test cases that check the changes thoroughly including negative and positive cases if possible

• Add screenshots for manual tests if appropriate

• Run test locally before make a pull request