I am trying to write column definitions (concat column name and data type) to a redshift table so that I can reuse these column defs when replaying archived s3 files for rep