Integrate snappy pull-data into workflow configuration

[xiamaz]

Is your feature request related to a problem? Please describe.
For most workflows data will need to be downloaded from SODAR and eventually reuploaded to SODAR. Currently the download step is fully independent of other workflow steps and mostly will need to be performed either through non-standard scripts or manually following project specific documentation.

Describe the solution you'd like
Instead of manually specifying input data sources, snappy configuration should contain required input filetypes for processing and all required filetypes should be pulled from SODAR.

Additional context
A potential solution might be extending the current snappy-pipeline config. Currently data_sets implements search_patterns and search_patterns, which can directly be leveraged for downloading data from SODAR. By adding input_filetypes pull-data can be extended to automatically download data for each SODAR repository into the respective target directory.