Re: [boost] lifetime of ranges vs. iterators

From: David Abrahams (dave_at_[hidden])
Date: 2008-08-31 15:19:11

on Sun Aug 31 2008, Arno Schödl <aschoedl-AT-think-cell.com> wrote:

>>  // Gimme a better name, please!  The abstraction is a thing that
>>  // is always used in pairs, where each instance contains some 
>>  // redundant information in common.
>>  auto concept Factorable<typename T>
>>  {
>>       typename common_data = void;
>>       typename unique_data = T;
>>
>>       common_data common(T) {}
>>       unique_data unique(T x) { return x; }
>>       T reconstruct(common_data, unique_data x) { return x; }
>>  };
>
> How is the separation of common_data and unique_data different from a separation
> of ranges and iterators? If iterators of ranges can rely on their range to
> exist, this is where common data like functors, end iterators etc. can be
> stored.

Naturally the point is to store the common data once when you need to
store more than one iterator to the same sequence -- in a range, for
example.

If you're asking why I bother reconstituting the whole iterator out of
its parts, instead of simply referring to the common_data stored in the
range: an adapted iterator is a useful concept regardless of whether
anyone is using a range abstraction.

-- 
Dave Abrahams
BoostPro Computing
http://www.boostpro.com