On Fri, Jul 19, 2013 at 07:26:09AM +0200, Michal Privoznik wrote:
On 19.07.2013 04:04, Doug Goldstein wrote:
> On Thu, Jul 18, 2013 at 1:13 PM, Dave Allan <dallan(a)redhat.com> wrote:
>> I've seen a bunch of interest in python3 [1,2,3]. Has anybody started
>> thinking about python3 bindings for libvirt?
>>
>
> While not really answering your question, I would like to see the
> build system a little more flexible with regards to Python binding
> generation. Currently you have to configure libvirt and it will
> generate bindings for the python it detects (or is supplied) and if
> you want to change that you need to rebuild the entire source tree.
> It'd be nice in the future to be able to build against multiple
> Python's without having to reconfigure and rebuild.
That's because python bindings need the client implementation. But I
agree that it would be nice if one could just:
make -C python
to build the bindings.
To answer Dave's question, I'm not really into python but isn't 2to3 enough?
2to3 will take a python2 file & spit out a reasonable python3
file. The issue is that, IMHO, we don't want to be in the business
of maintaining 2 sets of python bindings. For any C code, I think
we want to make sure we use #if conditionals to minimize the extra
burden of python3. For the py code, I think we should also aim to
try to support 2 & 3 with one set of code. There are some modules
you can import which give you access to some python3 style apis
from python2, which could help.
Daniel
--
|:
http://berrange.com -o-
http://www.flickr.com/photos/dberrange/ :|
|:
http://libvirt.org -o-
http://virt-manager.org :|
|:
http://autobuild.org -o-
http://search.cpan.org/~danberr/ :|
|:
http://entangle-photo.org -o-
http://live.gnome.org/gtk-vnc :|