S3-Parquet destination connector enhancements to use the actual schema data type instead of varchar for different data type columns

[hokam]

Currently we just use the varchar as data type for different type of source schema columns. Once the implementation is done the S3 parquet file and Athena table should have actual source data types.