[C++-sig] copy.deepcopy of a vector, and the resulting error
matthew.scouten at gmail.com
Fri Feb 8 18:09:19 CET 2008
If you are looking for agreement, you have mine. This seems like an obvious
good idea to me. Ideally, the person using my library should not have to
know what language I used to write it. Lacking deepcopy is an unacceptable
But I do have a few more questions,
1) This would be in the cvs only at first (I assume), so:
2) Will the latest cvs BP still be compatible with boost 1.34.1, or will I
need latest cvs for all of boost?
I need latest cvs for all of boost, that is a killer.
I would love to see this in BP, it will save me a lot of work.
On Feb 8, 2008 3:23 AM, Hans Meine <meine at informatik.uni-hamburg.de> wrote:
> Am Donnerstag, 07. Februar 2008 20:23:54 schrieb Matthew Scouten:
> > Implementing an explicit __deepcopy__ for every class could be a lot of
> > work , and I do not understand what that memo argument is doing.
> You can use my implementation.
> > I had hoped that BP would just know what to do. After all, all the
> > in question have perfectly good copy ctors. I take it that
> > 'class_<...>.enable_copy()' and the other ideas in that thread were
> > implemented?
> Yes, I have sent code that contains the implementation (in that thread,
> 2007-10-17 10:25), only:
> 1) The API is not class_<...>.enable_copy()
> 2) It was not integrated into the BPL (yet)
> AFAICS, David agreed in principle that this would be good to have in
> boost::python, and he even suggested to have that by default for
> not-noncopyable classes.
> If there is agreement on that, I will have a look at that in ca. 2 weeks;
> maybe I can propose a patch.
> Ciao, / /
> / / ANS
> C++-sig mailing list
> C++-sig at python.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Cplusplus-sig