OO, complexity and toons...

[Kragen Sitaker]

shagshag13 at yahoo.fr (Shagshag) writes:
> I agree with you, it's more about relational databases than
> oo-programming, but i really need to work on this kind of data
> structures (there is something about studying inverted files/index and
> algorithms...)
> 
> And, as i'm a real newbie in python and OOP, i would like to be sure
> that python could handle this amount of data, and still be quite
> effective in search.
> 
> So do you think my first idea is viable ? How, if you had to, will you
> design it in OO style ?

So your question is, "How do I implement a scalable RDBMS in Python,
if that's even possible?"  You could look at Gadfly to see whether
it's possible to do this in-memory.

I haven't done this kind of stuff myself, but I don't think it should
be more difficult in Python than in other languages, and my uninformed
guess is that you should be able to get quite acceptable performance
for the dataset sizes you were discussing.  Certainly you should be
able to get performance on modern machines in pure Python that rivals
what you could get on 1990 machines in assembler or C.