Automatically detect use_cache initial value
The base worker should now be able to detect if use_cache
is True
(we stick to False
by default).
BaseWorker & ElementsWorker should not have a use_cache
in their constructor, but a support_cache=False
instead.
Then self.use_cache
becomes True
, if (and only if):
-
support_cache
is True - an sqlite database is available in the parent tasks
The ML workers would then just have to declare if they support cached elements alongside normal elements