Structure of configuration file generated by BL

When you run an App locally, you use a config.json.example (to test your App), or if you're not the developer of the App, you use a config.json that has the same keys as the config.json.example available on the GitHub repository.

The keys of these configuration files are the input files corresponding to the datatype used and the parameters of the functions used by the App. For instance, for an App with a input neuro/meg/fif datatype where four optional files are listed in the file mapping:

{
    "fif": "/network/lustre/iss01/home/aurore.bussalb/Repositories/test_snr/data/subject_1/bad_channels_1.fif",
    "calibration": "/network/lustre/iss01/home/aurore.bussalb/Repositories/data_test_BL/sss_cal.dat",
    "crosstalk": "/network/lustre/iss01/home/aurore.bussalb/Repositories/data_test_BL/ct_sparse.fif",
    "destination": "/network/lustre/iss01/home/aurore.bussalb/Repositories/data_test_BL/mean_tm-raw.fif",
    "headshape": ""
}

However, in the config.json other keys are present in this file: _app, tid, _inputs, and _outputs.

The config file is parsed by the Python file of the App and becomes a Python dictionary. If you want to use this dictionary as a **kwargs for instance, you need to delete the keys generated by BL:

if '_app' and '_tid' and '_inputs' and '_outputs' in config.keys():
    del config['_app'], config['_tid'], config['_inputs'], config['_outputs'] 
kwargs = config