When generating forcing from caravan, the entire caravan shapefile (100 MB) is downloaded and unzipped each time, unless a path to the shapefile of the individual basin is provided. Timing issues between python and OS (I think?) cause this to fail when multiple catchments are requested in rapid succession.
As a researcher who wants to study the entire camels dataset for my research, I want to be able to point to a location where the combined shapefile is stored (cached) to avoid this overhead.
A solution would be to check if the current optional "shape" argument points to a file called "combined.shp". If it does, it is assumed this is the combined caravan shapefile and the shape of the selected catchment is extracted from it.