Currently when using the CSV Buffer to export data to a csv, the WriteToStream procedure expects all lines in the buffer to have the same number of columns.

The Idea would be to check the number of columns for each row before writing the contents of the CSVBuffer to the StreamWriter. This could be stored in a dictionary and then checked before deciding if a FileSeparator is to be added or new line taken.

For OnPrem sites we have written bespoke code on the table extension to get around this issue but as the StreamWriter DotNet variable cannot be accessed when using Cloud, a solution is needed in the Base App.