Enhance concept of extension of job

[eileen-kuehn]

In the context of caching the Job is defined by further information such as inputfiles but also statistics that are relevant for caching. The current implementation provides an extension of the basic Job and implements a CachingJob, see lapis/cachingjob.py.
With further extensions we might face issues by subclassing and specifically by overwriting the run method as it carries relevant logic for the simulation that is currently copied from the core Job:

lapis.caching/lapis/cachingjob.py

Lines 207 to 246 in 8f4f614

	async def run(self, drone: "Drone"):
	"""
	Handles the job's execution.
	The job's runtime is given by max(calculation time, transfer time).
	The calculation time is determined by `_calculate`, the transfer time by
	`_transfer_inputfiles`.
	The job will be executed successfully unless the selected drone does not
	provide enough resources, is unavailable or an exception occurs.
	:param drone: the drone object the job was allocated to and is executed in
	"""
	assert drone, "Jobs cannot run without a drone being assigned"
	self.drone = drone
	self.in_queue_until = time.now
	self._success = None
	await sampling_required.put(self)
	try:
	start = time.now
	print(start)
	async with Scope() as scope:
	await instant
	scope.do(self._transfer_inputfiles())
	await (time + self._calculation_time)
	except CancelTask:
	print("CancelTask")
	# self.drone = None
	self._success = False
	# await sampling_required.put(self)
	# TODO: in_queue_until is still set
	except BaseException:
	# self.drone = None
	self._success = False
	await sampling_required.put(self)
	# TODO: in_queue_until is still set
	raise
	else:
	self.walltime = time.now - start
	self._success = True
	await sampling_required.put(self)

For future maintainability we should consider a different approach here.