huge persistent dictionary

[Cameron Laird]

In article <slrn90gg2u.d3s.quinn at spew.ugcs.caltech.edu>,
Quinn Dunkan <quinn at spew.ugcs.caltech.edu> wrote:
>On Tue, 07 Nov 2000 13:05:10 -0000, gunter  <gb at a3design.de> wrote:
>>Hello!
>>
>>My Problem is, that I want to store an huge amount of data in an
>>dictionary and to make it persistent.
>...
>Others have made some good suggestions, but I'd recommend you try marshal
>first.  It only works on primitive types, but it's faster than cPickle.  And
>it's essentially the same as your import trick (it writes and reads python
>bytecode), delta the huge intermediate file.
>
>still-waiting-for-someone-to-post-a-mmap-solution-ly y'rs

At least four people have said, "Use the D database", for four
different values of D.  As Mr. Dunkan solicited a mmap() solution,
I'll counter with MetaKit <URL:http://www.equi4.com/metakit/#h3>.
While less widely known than the other alternatives suggested, it
tends to inspire wild enthusiasm in its users.  Salient features:
it's Python-friendly, memory-based, documented as handling gigabytes
with high performance, and well-maintained.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html