This is achievable by carefully crafting a callback to submit the second operation after the first one has completed. Sadly, it is not possible to pass an arbitrary future to pool.submit
so an extra step is required to bind the two futures together.
Here is a possible implementation:
import concurrent.futures
def copy_future_state(source, destination):
if source.cancelled():
destination.cancel()
if not destination.set_running_or_notify_cancel():
return
exception = source.exception()
if exception is not None:
destination.set_exception(exception)
else:
result = source.result()
destination.set_result(result)
def chain(pool, future, fn):
result = concurrent.futures.Future()
def callback(_):
try:
temp = pool.submit(fn, future.result())
copy = lambda _: copy_future_state(temp, result)
temp.add_done_callback(copy)
except:
result.cancel()
raise
future.add_done_callback(callback)
return result
Note that copy_future_state
is a slightly modified version of asyncio.futures._set_concurrent_future_state.
Usage:
from concurrent.futures import ProcessPoolExecutor
def wait(seconds):
time.sleep(seconds)
return seconds
pool = ProcessPoolExecutor()
future1 = pool.submit(wait, 5)
future2 = chain(pool, future1, wait)
future3 = pool.submit(wait, 10)