I am trying to output CSV files to a specific directory structure like so:
directory1/
bike/
ride_log.csv
bike_features.csv
directory2/
bike/
customers.csv
When I run the Data Download job the resulting file structure in the output zip file looks like this:
directory1/
bike/
ride_log.csv
bike_features.csv
directory2/
2/
customers.csv
The CSV features ride_log and bike_features belong to the same CSV writer: bike_1, and the customers feature belongs to a CSV writer: bike_2.
The bike_1 and bike_2 writers have Dataset Fanout applied based on an attribute named 'directory' which gets set to 'bike' in the transformation process of all three features. There is also a prefix directory set for bike_1 as 'directory1', and a prefix directory for bike_2 as 'directory2'. Both features have their output directory linked to $FME_SERVER_DEST_DIR, which as an FME server parameter is set to \.
So how can I get the same folder name in two different subdirectories when using fanout like this, or why is FME overwriting one of the fanout directory names when they are the same even though they come from different writers and have different directory prefixes assigned.