currently it's quite basic, there are a few points that could be improved: * more control over the format of the output * file names should be based on resource names (need slugify + detect duplicates because resource names supports unicode)