- Boost - lists.boost.org

[multi_index] Another weird problem with MI on MSVC 7.1
by Alexei Alexandrov 26 May '06

26 May '06

Hi, I have a problem compiling the code snippet below with MSVC 7.1 (Intel Compiler 9.0 and GCC compile it just fine). The problem trigger is keyword "virtual" in class base inheritance - if you remove it it will compile OK. The error message is very weird: boost\type_traits\is_function.hpp(65) : error C2371: 'boost::detail::t' : redefinition; different basic types boost\type_traits\is_function.hpp(65) : see declaration of 'boost::detail::t' boost\type_traits\is_function.hpp(82) : see reference to class template instantiation '<Unknown>' being compiled boost\tuple\detail\tuple_basic.hpp(310) : see reference to class template instantiation '<Unknown>' being compiled boost\tuple\detail\tuple_basic.hpp(418) : see reference to class template instantiation '<Unknown>' being compiled boost\multi_index\detail\def_ctor_tuple_cons.hpp(33) : see reference to class template instantiation '<Unknown>' being compiled boost\multi_index_container.hpp(140) : see reference to class template instantiation '<Unknown>' being compiled Also, it is critical for reproducing to have 2 independent inheritance chains - if I remove name_record_set2 class and variable it also compiles just fine. Here is the code itself: #include <boost/multi_index_container.hpp> #include <boost/multi_index/mem_fun.hpp> #include <boost/multi_index/ordered_index.hpp> using namespace boost::multi_index; class base {}; class base_record : virtual public base { public: base_record() : id_(0) {} int get_id() const { return id_; } private: int id_; }; class name_record : public base_record {}; class name_record2 : public base_record {}; typedef multi_index_container< name_record, indexed_by< ordered_unique< BOOST_MULTI_INDEX_CONST_MEM_FUN(base_record, int, get_id) > > > name_record_set; typedef multi_index_container< name_record2, indexed_by< ordered_unique< BOOST_MULTI_INDEX_CONST_MEM_FUN(base_record, int, get_id) > > > name_record_set2; int main() { name_record_set ns; name_record_set2 ns2; ns.insert(name_record()); ns2.insert(name_record2()); return 0; } P.S. I use boost from RC_1_34 tag. -- Alexei Alexandrov

3 3

Re: [boost] Infinite precision integer draft
by Martin Bonner 26 May '06

26 May '06

----Original Message---- From: boost-bounces(a)lists.boost.org [mailto:boost-bounces@lists.boost.org] On Behalf Of Maarten Kronenburg Sent: 26 May 2006 13:09 To: boost(a)lists.boost.org Subject: Re: [boost] Infinite precision integer draft > const integer get_sub( const integer &, const integer & ) const; > By the way here no size_type is used, because > size_type suffices for the range of bytes, > but not for the range of bits, which is a factor > 8 larger. Is that worth worrying about? Even on 32-bit operating systems, a 4 gigabit integer is pretty big. -- Martin Bonner Martin.Bonner(a)Pitechnology.com Pi Technology, Milton Hall, Ely Road, Milton, Cambridge, CB4 6WZ, ENGLAND Tel: +44 (0)1223 203894

3 2

[format] bug and proposed patch
by Joaquín Mª López Muñoz 26 May '06

26 May '06

Edson Tadeu reports on the following post http://lists.boost.org/boost-users/2006/05/19723.php about this bug in Boost.Format: line 138 of boost/format/format_implementation.hpp if( bound_.size()==0 || !bound_[ items_[i].argN_ ] ) can try to invalidly evaluate bound_[...] on vector bound_ with a negative index, which happens when items_[i].argN_ is one of the special values enum arg_values { argN_no_posit = -1, // non-positional directive. will set argN later argN_tabulation = -2, // tabulation directive. (no argument read) argN_ignored = -3 // ignored directive. (no argument read) }; as defined in boost/format/internals.hpp. The attached patches to boost/format/format_implementation.hpp and libs/format/test/format_test3.cpp avoid the invalid negative indexing and add a testcase for the problem, respectively. Any objections to my commiting? If there's no comment I'll be doing it tomorrow. Thank you, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo 138c138 < if( bound_.size()==0 || !bound_[ items_[i].argN_ ] ) --- > if( bound_.size()==0 || items_[i].argN_<0 || !bound_[ items_[i].argN_ ] ) 95c95,102 < --- > > // testcase for bug reported at > // http://lists.boost.org/boost-users/2006/05/19723.php > format f("%40t%1%"); > int x = 0; > f.bind_arg(1, x); > f.clear(); >

1 1

[array] optional constructors in Boost.Array
by Vincent Bherer-Roy 26 May '06

26 May '06

I know that this subject has been discuss a lot in the past. However a quick search in the mailing list archive did not show any solution like the one I propose. If my post is a duplicate, please forgive :-) The problem is that, like many other, I don't use boost::array because it doesn't have any constructor. First, be aware that I want to start a flame war ;-) I understand well the rationale for the aggregate initialization support. I also understand that we want a lightweight wrapper around C arrays. I agree there is a consensus around this and I don't want to break it. However, I have a simple (I think) solution that should give us the best of both worlds. Basically, the idea is to have a third parameter to the template, it is set to false by default in order to provide the same default behaviour. So we have: template<class T, std::size_t N, bool C = false> class array { /* current implementation */ }; template<class T, std::size_t N> class array<T, N, true> : public array<T,N,false> { /* adds constructors only */ }; array<int, 3> a1 = { 1, 2, 3 }; array<int, 3, false> a2 = { 1, 2, 3 }; array<int, 3, true> a3; array<int, 3, true> a4( 1 ); // assign 1 to all three values array<int, 3, true> a5( 1, 2, 3 ); array<int, 3, true> a6( 1, 2 ); // raises a static assert because we try to put only two values in an array of 3 array<int, 3, true> a7( 1, 2, 3, 4 ); // also raises a static assert because we try to put four values in an array of 3 I only see one main drawback with this technique, we'll have some problems with compilers lacking partial specialization support. So the idea is in your hands. I hope it will have some echoes. I attached the modified array.hpp file. It compiles for me on Visual Studio .Net 2003 and 2005 with boost 1.33.1. However I didn't make full test coverage. Thanks Vincent Bherer-Roy

5 13

Re: [boost] Infinite precision integer draft
by Martin Bonner 26 May '06

26 May '06

Gerhard Wesp Sent: > On Thu, May 25, 2006 at 04:41:32PM +0200, Maarten Kronenburg wrote: >> Users can implement an unsigned_integer by deriving from >> integer, but this will not give any performance gain. > > This would violate Item 32 in Effective C++, 3rd ed. by Scott Meyers: > Make sure public inheritance models "is-a". I agree it's not a good idea. > I suggest deriving from integer be disallowed in the TR2 proposal. ... however I don't think that is. If C++ had "final", then it could be compile time checked. As it is, I don't think TR2 could do more than make it undefined behaviour. We really don't want more undefined behaviour in C++ if we can avoid it. What is more, deriving from integer would actually work on all implementations (as far as public inheritance which doesn't model "is-a" works). -- Martin Bonner Martin.Bonner(a)Pitechnology.com Pi Technology, Milton Hall, Ely Road, Milton, Cambridge, CB4 6WZ, ENGLAND Tel: +44 (0)1223 203894

2 1

[review][pimpl_ptr] Pimpl Pointer review begins
by Rene Rivera 26 May '06

26 May '06

The review of the pimpl_ptr library by Asger Mangaard starts today, Monday May 15th, 2006, and runs through May 25th, 2006. (Boost Reviews RSS feed <http://tinyurl.com/ogpj6>) Pimpl Pointer ------------- :Author: Asger Mangaard :Download: Boost Sandbox (http://boost-consulting.com/vault/) as "pimpl_ptr.zip". Direct download at <http://tinyurl.com/m66ox>. :Description: The pimpl idiom is widely used to reduce compile times and disable code coupling. It does so by moving private parts of a class from the .hpp file to the .cpp file. However, it's implementation can be tricky, and with many pitfalls (especially regarding memory management). The pimpl_ptr library is a single header file, implementing a special policy based smart pointer to greatly ease the implementation of the pimpl idiom. Review questions ---------------- Please always explicitly state in your review, whether you think the library should be accepted into Boost. You might want to comment on the following questions: - What is your evaluation of the design? - What is your evaluation of the implementation? - What is your evaluation of the documentation? - What is your evaluation of the potential usefulness of the library? - Did you try to use the library? With what compiler? Did you have any problems? - How much effort did you put into your evaluation? A glance? A quick reading? In-depth study? - Are you knowledgeable about the problem domain? Enjoy, Rene. -- -- Grafik - Don't Assume Anything -- Redshift Software, Inc. - http://redshift-software.com -- rrivera/acm.org - grafik/redshift-software.com -- 102708583/icq - grafikrobot/aim - grafikrobot/yahoo

12 28

Anyone Using TortoiseCVS with anonymous to Checkout/update Boost from Sourceforge?
by Paul A Bristow 25 May '06

25 May '06

I am failing to get TortoiseCVS to checkout or update Boost files at Sourceforge. Since I don't have an account, I am trying to use Username: anonymous (This works OK at present using comandline CVS, using same connection - Firewall etc). I get :-((( "In H:\: "C:\Program Files\TortoiseCVS\cvs.exe" "-z3" "checkout" "-P" "boost/libs/math/test" CVSROOT=:pserver:anonymous@boost-sandbox.cvs.sourceforge.net:/cvsroot/boost- sandbox connect to boost-sandbox.cvs.sourceforge.net:2401 failed: A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond. Success, CVS operation completed" Is anyone using this sucessfully like this (anonymous) Or can anyone suggest what I am doing wrong. Thanks Paul --- Paul A Bristow Prizet Farmhouse, Kendal, Cumbria UK LA8 8AB +44 1539561830 & SMS, Mobile +44 7714 330204 & SMS pbristow(a)hetp.u-net.com

2 1

Additions to optional<>
by Fernando Cacciola 25 May '06

25 May '06

Hello I have a couple of pending additions to the Optional library. These are _new_ functions, so they shouldn't disrupt the release. Specifically: from_optional(o,default_value) and make_optional(condition,value) Given that the release is still ahead, is it OK to merge these additions to RC_1_34_0 or should I keep them in the trunk? TIA -- Fernando Cacciola SciSoft http://fcacciola.50webs.com/

3 2

[review][pimpl_ptr] Review ends today.
by Rene Rivera 25 May '06

25 May '06

This is a reminder that the review period for the pimpl_ptr library by Asger Mangaard ends today. If you have a few minutes to do a review please do so ASAP ;-) <http://lists.boost.org/Archives/boost/2006/05/104757.php> -- -- Grafik - Don't Assume Anything -- Redshift Software, Inc. - http://redshift-software.com -- rrivera/acm.org - grafik/redshift-software.com -- 102708583/icq - grafikrobot/aim - grafikrobot/yahoo

1 0

Re: [boost] [array] optional constructors in Boost.Array
by Martin Bonner 25 May '06

25 May '06

----Original Message---- From: boost-bounces(a)lists.boost.org [mailto:boost-bounces@lists.boost.org] On Behalf Of Thorsten Ottosen Sent: 25 May 2006 15:20 To: boost(a)lists.boost.org Subject: Re: [boost] [array] optional constructors in Boost.Array > Martin Bonner wrote: >> Thorsten Ottosen wrote: >> >>> Maybe you should just wrap boost::array: >>> >>> template< class T, unsigned N > >>> struct my_array : boost::array<T,N> >>> { >>> // provide constructors here >> >> >> How? >> >> >>> }; >> >> >> >> You can't construct the base boost_array in the base class >> initialization list of the constructor (because boost_array doesn't >> have a constructor that takes arguments), and by the time you get >> into the body of the my_array constructor it is too late - the >> boost_array is already default constructed. > > What do tou mean by "it's too late"? What I mean is, you can't construct the elements of the boost_array at that point. > > Since the "construction" of boost::array is a no-op, you can just > initialize it in the body my_array's contructors. I guess it depends what you want to use my_array for. If you are trying to hold primitive types, then that wrapping works fine. If you are trying to hold objects of a class type that doesn't even /have/ a default constructor, then it just won't work! (and yes, you can put such objects in a boost_array on a compliant compiler). -- Martin Bonner Martin.Bonner(a)Pitechnology.com Pi Technology, Milton Hall, Ely Road, Milton, Cambridge, CB4 6WZ, ENGLAND Tel: +44 (0)1223 203894

3 2