Hi. Is it any difference in task calling syntax?

class SetupStorageTask(Task): 
    def run(self, mode) -> Storage:    
        return Storage(mode=mode)

setup_storage_task = SetupStorageTask()
storage = setup_storage_task(mode='rw')

and

class SetupStorageTask(Task):

    def __init__(self, mode):
        self.mode = mode
    
    def run(self) -> Storage:
        return Storage(mode=self.mode)

setup_storage_task = SetupStorageTask(mode='rw')
storage = setup_storage_task()

or can be just
storage = SetupStorageTask(mode='rw')

Both can work, it depends on how you want to use this task. E.g. if you want to pass some data into the task, then it makes sense to accept this argument within the “run” method rather than only on init. Many tasks in our task library set all attributes as optional allowing them to be set both at initialization and within the Flow constructor and include some logic within the task itself to check whether the required values are set and if not raise ValueError. I think if you have a look at a couple of tasks in the task library code, you’ll get a better grasp of how it works.

no, it is project related storage

It has to be :

storage = SetupStorageTask(mode='rw')()

The first parenthesis is the

init

and the second is the

run

but it works without last ()

yes, you can. Again, I think you’ll get a better feel for it when you look at some existing tasks, there are many that use a similar pattern but if you’re asking how it should be called within Flow, then as Kevin said:

with Flow() as flow:
   storage = SetupStorageTask(mode='rw')()

It makes sense I just was surprised that without last () it do the same thing