On Fri, Jun 26, 2009 at 1:52 PM, Simon King<[email protected]> wrote:
>
> Dear Lars,
>
> On Jun 26, 1:43 pm, Lars Fischer <[email protected]>
> wrote:
>> you could try:
>> import pGroupCohomology.mtx  as mtx
>
> This is one of the things that I tried, but it didn't help. When
> loading a pickle of some class instance from mtx, it was still
> complaining.
>
> Probably, in the long run, the easiest is to temporarily have a fake
> mtx module on top level, that imports the necessary classes from
> pGroupCohomology.mtx. Then,
>  1. load the old data: Loading looks for mtx.SomeClass, and this
> results in pGroupCohomology.mtx.SomeClass.
>  2. save the data back. Now, the pickle points to
> pGroupCohomology.mtx.SomeClass
>  3. remove the fake mtx, in order to not pollute the top level
>  4. load the data. This is possible since the "address" of "SomeClass"
> was automatically corrected when saving the data back.
>
> I know that this procedure works, but I was wondering if something
> simpler is possible.

We do the above in several places in the Sage library right now.
There is something else that is possible though -- I recall Craig
Citro and Carl Witty discussing it -- but I can't remember exactly
what it is.

William

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at 
http://groups.google.com/group/sage-support
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

Reply via email to