Read/write files from Amazon S3 bucket

Some GCW contributors have their raw data in an AWS S3 bucket. For GCW, it would be nice to support this use case, knowing that individual files can be retrieved with curl (also mounted?). See AWS documentation.

The core question could be at which level and how we implement such a feature: a new data transport layer in MeteoIO (an old dream of mine)? Some kind of mount in the container of the web service? Synchronization by curl in the web service container?

For Envidat, R/W support is highly recommended (Users can also easily share their data on S3 and Envidat can link to final published datasets on S3).