Add new column while converting Json to Parquet

Question

Json to parquet conversion is working fine, but I need to add a new column in parquet table containing timestamp so that I can query records in latest to oldest order.

Currently I am using the following script:

def toString(rec):
    for key in rec:
        rec[key] = str(rec[key])
    return rec

def convertIntoParquet(sourceTableName):
    # relationalize the json schema.
    datasource = glueContext.create_dynamic_frame.from_catalog(database = glueSourceDatabase, table_name = sourceTableName, transformation_ctx = "datasource")
    relationalizedDatasource = Relationalize.apply(frame = datasource, staging_path = glueTempStorage, name = jsonRootTableName, transformation_ctx = "relationalizedDatasource")
    transactionData = relationalizedDatasource.select(jsonRootTableName)
    modifiedTransactionData = Map.apply(frame = transactionData, f = toString)

    # adding new partitions
    modifiedTransactionData_df = modifiedTransactionData.toDF()
    print("Total transactions in dataframe : " , modifiedTransactionData_df.count())
    if(modifiedTransactionData_df.count() == 0):
        return

    modifiedTransactionData_df = modifiedTransactionData_df.repartition(col("panHash"), col("financialYear"), col("accountId"))

    modifiedTransactionData = DynamicFrame.fromDF(modifiedTransactionData_df, glueContext, "modifiedTransactionData")

    # converts in parquet format.
    sink = glueContext.getSink(
        connection_type = "s3",
        path = glueRelationalizeOutputS3Path,
        partitionKeys = ["panHash", "financialYear", "accountId"],
        transformation_ctx = "sink")
    sink.setFormat("glueparquet")
    sink.writeFrame(modifiedTransactionData)