On Wed, Feb 12, 2014 at 3:10 PM, Ethan Furman <et...@stoneleaf.us> wrote:
> On 02/12/2014 02:02 PM, yury.selivanov wrote: > >> http://hg.python.org/cpython/rev/920304e1f36b >> changeset: 89175:920304e1f36b >> user: Yury Selivanov <yseliva...@sprymix.com> >> date: Wed Feb 12 17:01:52 2014 -0500 >> summary: >> asyncio.events: Use __slots__ in Handle and TimerHandle >> >> files: >> Lib/asyncio/events.py | 4 ++++ >> 1 files changed, 4 insertions(+), 0 deletions(-) >> >> >> diff --git a/Lib/asyncio/events.py b/Lib/asyncio/events.py >> --- a/Lib/asyncio/events.py >> +++ b/Lib/asyncio/events.py >> @@ -19,6 +19,8 @@ >> class Handle: >> """Object returned by callback registration methods.""" >> >> + __slots__ = ['_callback', '_args', '_cancelled'] >> + >> def __init__(self, callback, args): >> assert not isinstance(callback, Handle), 'A Handle is not a >> callback' >> self._callback = callback >> @@ -46,6 +48,8 @@ >> class TimerHandle(Handle): >> """Object returned by timed callback registration methods.""" >> >> + __slots__ = ['_when'] >> + >> def __init__(self, when, callback, args): >> assert when is not None >> super().__init__(callback, args) >> > > Apologies up front if these are stupid questions, but: > > Why __slots__? Are we going to have so many of these things that memory > is an issue > There's one of these created for every callback -- which includes every time 'yield from' blocks for a Future. The savings are primarily in allocation cost (no dicts to allocate). > The asserts in the code -- those are not for checking user input, correct? > Yeah, this class is never instantiated directly by the user. -- --Guido van Rossum (python.org/~guido)
_______________________________________________ Python-Dev mailing list Python-Dev@python.org https://mail.python.org/mailman/listinfo/python-dev Unsubscribe: https://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com
