<p dir="ltr">Funny, I suggested these 2 in the past:<br>
<a href="https://groups.google.com/forum/m/#!searchin/python-ideas/map_as_completed/python-ideas/VZBdUbYcQjg">https://groups.google.com/forum/m/#!searchin/python-ideas/map_as_completed/python-ideas/</a><a href="https://groups.google.com/forum/m/#!searchin/python-ideas/map_as_completed/python-ideas/VZBdUbYcQjg">VZBdUbYcQjg</a></p>
<p dir="ltr"><a href="https://groups.google.com/forum/m/#!searchin/python-ideas/as_completed/python-ideas/yGADxChihhk">https://groups.google.com/forum/m/#!searchin/python-ideas/as_completed/python-ideas/yGADxChihhk</a></p>
<p dir="ltr">Sent from my phone.</p>
<div class="gmail_quot<blockquote class=" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">On 2 May 2015 at 19:25, Ram Rachum <<a href="mailto:ram@rachum.com">ram@rachum.com</a>> wrote:<br>
> Okay, I implemented it. Might be getting something wrong because I've never<br>
> worked with the internals of this module before.<br>
<br>
I think this is sufficiently tricky to get right that it's worth<br>
adding filter() as a parallel to the existing map() API.<br>
<br>
However, it did raise a separate question for me: is it currently<br>
possible to use Executor.map() and the as_completed() module level<br>
function together? Unless I'm missing something, it doesn't look like<br>
it, as map() hides the futures from the caller, so you only have<br>
something to pass to as_completed() if you invoke submit() directly.<br>
<br>
Cheers,<br>
Nick.<br>
<br>
--<br>
Nick Coghlan   |   <a href="mailto:ncoghlan@gmail.com">ncoghlan@gmail.com</a>   |   Brisbane, Australia<br>
</div>