multiprocessing.Condition.wait_for missing

[sbt]

BPO	14087
Nosy	@pitrou
Files	cond_wait_for.patch cond_wait_for.patch cond_wait_for.patch cond_wait_for.patch

^{Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.}

Show more details

GitHub fields:

assignee = None
closed_at = <Date 2012-04-17.17:30:15.777>
created_at = <Date 2012-02-22.15:01:24.967>
labels = ['type-feature', 'library']
title = 'multiprocessing.Condition.wait_for missing'
updated_at = <Date 2012-04-17.17:30:15.775>
user = 'https://bugs.python.org/sbt'

bugs.python.org fields:

activity = <Date 2012-04-17.17:30:15.775>
actor = 'neologix'
assignee = 'none'
closed = True
closed_date = <Date 2012-04-17.17:30:15.777>
closer = 'neologix'
components = ['Library (Lib)']
creation = <Date 2012-02-22.15:01:24.967>
creator = 'sbt'
dependencies = []
files = ['24611', '24626', '25155', '25240']
hgrepos = []
issue_num = 14087
keywords = ['patch']
message_count = 11.0
messages = ['153956', '154063', '154093', '154135', '157244', '157762', '158459', '158497', '158533', '158557', '158559']
nosy_count = 4.0
nosy_names = ['pitrou', 'neologix', 'python-dev', 'sbt']
pr_nums = []
priority = 'normal'
resolution = 'fixed'
stage = 'resolved'
status = 'closed'
superseder = None
type = 'enhancement'
url = 'https://bugs.python.org/issue14087'
versions = ['Python 3.3']

[sbt]

multiprocessing.Condition is missing a counterpart for the wait_for() method added to threading.Condition in Python 3.2.

I will work on a patch.

[sbt]

Patch which just copies the implementation from threading.

[neologix]

In test/test_multiprocessing.py:
"""
def test_waitfor(self):
[...]
with cond:
result = cond.wait_for(lambda : state.value==0)
self.assertTrue(result)
self.assertEqual(state.value, 0)
for i in range(4):
time.sleep(0.01)
with cond:
state.value += 1
cond.notify()
"""

Shouldn't the for loop be outside the outer with block?

In Lib/multiprocessing/managers.py:
Is there a good reason why the wait_for() proxy method can't simply be implemented as:
return self._callmethod('wait_for', (predicate, timeout))?

(There may be, I just didn't check).

Finally, the documentation should be updated (Doc/library/multiprocessing.rst).

Otherwise, it would probably be better if you could submit a contributor agreement (and also maybe update your name on the tracker), unless that's a problem for you?

[sbt]

Shouldn't the for loop be outside the outer with block?

Yes.

In Lib/multiprocessing/managers.py:
Is there a good reason why the wait_for() proxy method can't simply be
implemented as:
return self._callmethod('wait_for', (predicate, timeout))?

(There may be, I just didn't check).

That would only work if predicate is picklable, which is unlikely to be the case. (The lambda functions used in the unit tests are not picklable.)

Finally, the documentation should be updated
(Doc/library/multiprocessing.rst).

Fixed in new patch.

Otherwise, it would probably be better if you could submit a contributor
agreement (and also maybe update your name on the tracker), unless that's
a problem for you?

I will try to submit it over the weekend.

[pitrou]

AFAICT (Raw)Value relies on ctypes, so the tests should be skipped if ctypes is unavailable. Or I guess you could use a Semaphore instead.

[sbt]

New patch skips tests if ctypes not available.

[sbt]

New patch which calculates endtime outside loop.

[pitrou]

Charles-François, will you take this one? :)

[neologix]

Charles-François, will you take this one? :)

Yes :-)

[python-dev]

New changeset 5606ee052783 by Charles-François Natali in branch 'default':
Issue bpo-14087: multiprocessing: add Condition.wait_for(). Patch by sbt.
http://hg.python.org/cpython/rev/5606ee052783

[neologix]

Committed.
Thanks for the patch, and sorry for the delay!