segemented iterator support?
Does boost support a kind of segemented iterator (Segmented Iterators and Hierarchical Algorithms, Matthew H. Austern)? Oliver
k-oli@gmx.de wrote:
Does boost support a kind of segemented iterator (Segmented Iterators and Hierarchical Algorithms, Matthew H. Austern)?
No. There isn't a library of algorithms in Boost. If there were, it might be advantageous to have hierarchical forms for some, but these would need to be privy to the internal details of the iterators passed to it, which wouldn't be the case in general. -- Eric Niebler BoostPro Computing http://www.boostpro.com
on Wed Dec 31 2008, Eric Niebler <eric-AT-boost-consulting.com> wrote:
k-oli@gmx.de wrote:
Does boost support a kind of segemented iterator (Segmented Iterators and Hierarchical Algorithms, Matthew H. Austern)?
No. There isn't a library of algorithms in Boost. If there were, it might be advantageous to have hierarchical forms for some, but these would need to be privy to the internal details of the iterators passed to it, which wouldn't be the case in general.
Well, I think that part about "privy to internal details" is a bit strong at best. They could be privy to a common generic interface for segmented iterators as suggested by Matt A. in his paper. You can get Matt's paper here: http://lafstern.org/matt/publications.html -- Dave Abrahams BoostPro Computing http://www.boostpro.com
David Abrahams wrote:
on Wed Dec 31 2008, Eric Niebler <eric-AT-boost-consulting.com> wrote:
k-oli@gmx.de wrote:
Does boost support a kind of segemented iterator (Segmented Iterators and Hierarchical Algorithms, Matthew H. Austern)?
No. There isn't a library of algorithms in Boost. If there were, it might be advantageous to have hierarchical forms for some, but these would need to be privy to the internal details of the iterators passed to it, which wouldn't be the case in general.
Well, I think that part about "privy to internal details" is a bit strong at best. They could be privy to a common generic interface for segmented iterators as suggested by Matt A. in his paper. You can get Matt's paper here: http://lafstern.org/matt/publications.html
I know. But many interesting segmented iterators will be from various STL implementations (e.g. std::deque::iterator). Who implements the customization points for those? And what if their implementation details are private? -- Eric Niebler BoostPro Computing http://www.boostpro.com
on Fri Jan 02 2009, Eric Niebler <eric-AT-boost-consulting.com> wrote:
David Abrahams wrote:
on Wed Dec 31 2008, Eric Niebler <eric-AT-boost-consulting.com> wrote:
k-oli@gmx.de wrote:
Does boost support a kind of segemented iterator (Segmented Iterators and Hierarchical Algorithms, Matthew H. Austern)?
No. There isn't a library of algorithms in Boost. If there were, it might be advantageous to have hierarchical forms for some, but these would need to be privy to the internal details of the iterators passed to it, which wouldn't be the case in general.
Well, I think that part about "privy to internal details" is a bit strong at best. They could be privy to a common generic interface for segmented iterators as suggested by Matt A. in his paper. You can get Matt's paper here: http://lafstern.org/matt/publications.html
I know. But many interesting segmented iterators will be from various STL implementations (e.g. std::deque::iterator). Who implements the customization points for those?
I guess we do, on a case-by-case basis.
And what if their implementation details are private?
Yes, that's an issue. -- Dave Abrahams BoostPro Computing http://www.boostpro.com
participants (3)
-
David Abrahams -
Eric Niebler -
k-oli＠gmx.de