Efficient implementation of deeply nested lists

[Bryan Olson]

Kay Schluehr wrote:
> I want to manipulate a deeply nested list in a generic way at a
> determined place. Given a sequence of constant objects a1, a2, ..., aN
> and a variable x. Now construct a list from them recursively:
> 
> L = [a1, [a2, [....[aN, [x]]...]]
> 
> The value of x is the only one to be changed. With each value of x a
> new instance of L is needed so that we actually have a function:
> 
> L = lambda x: [a1, [a2, [....[aN, [x]]...]]
> 
> I'm seeking for an efficient implementation that does not recompute the
> whole list each time. Any ideas?

The problem is that all your lists/sublists have different
values. "recompute" doesn't really even apply. If x != y,
then no two sublists found within L(x) plus L(y) are equal
(excluding sublists that might be within x or y).


-- 
--Bryan