- Boost-users - lists.boost.org

Re: [Boost-users] performance issue with graph edge removal
by Feldman, Yulik 25 Jul '05

25 Jul '05

I've also noticed that when in/out edge lists are stored in an associative container and the parallel edges are allowed, i.e multiset is used, the functions remove_edge()/clear_vertex() do not work as efficient as they could. The calls to these functions (in my case) result in the following sequence of calls: // O(E/V) template <class EdgeList, class vertex_descriptor> void erase_from_incidence_list(EdgeList& el, vertex_descriptor v, allow_parallel_edge_tag) { boost::graph_detail::erase_if(el, detail::target_is<vertex_descriptor>(v)); } Then: template <class Container, class Predicate> void erase_if(Container& c, Predicate p) { erase_if_dispatch(c, p, container_category(c), iterator_stability(c)); } Then: template <class AssociativeContainer, class Predicate> void erase_if_dispatch(AssociativeContainer& c, Predicate p, associative_container_tag, stable_tag) { typename AssociativeContainer::iterator i, next; for (i = next = c.begin(); next != c.end(); i = next) { ++next; if (p(*i)) c.erase(i); } } The complexity of the search operation is linear, so multiset doesn't provide a performance advantage over containers such as vector or set. However, the function erase_if_dispatch() could use multiset's equal_range() member function, which would result in O(log n) complexity of the search. I understand that the reason erase_if_dispatch() is written in the way it is written, is probably that the function is generic with regard to the predicate and doesn't do assumptions on it. However, in the case erase_if_dispatch() is called from erase_from_incidence_list(), where the predicate is set, the function (its special specialization) could make assumption that the predicate is actually the same as the predicate used to sort the in/out edge list itself (the target vertex), and therefore could use equal_range() to boost the performance of the function. --Yulik.

2 1

Re: [Boost-users] minimal spanning trees & Visual C++ 7.1
by Amit K. Jain 25 Jul '05

25 Jul '05

Thank you, it worked! It is bizarre behavior on the compiler's part, I hope they get it fixed eventually. -Amit > Hi! > > I suffered from ICEs using boost as well. They occured using multi_index lib, though. Turning of "Minimal Rebuild" helped for me. > > kind regards > > Alex

2 1

Missing functions in lexical_cast
by Sean Rohead 25 Jul '05

25 Jul '05

I would like to convert between std::wstring and std::string using lexical_cast and finally discovered in an old posting that you must manually provide template specializations for these types as below. My question is why these aren't provided in lexical_cast.hpp already? template<> inline std::wstring boost::lexical_cast(std::string source) { std::wstring result(source.size(), wchar_t(0)); typedef std::ctype<wchar_t> ctype_t; const ctype_t& ct = std::use_facet<ctype_t>(std::locale()); ct.widen(source.data(), source.data() + source.size(), &(*result.begin())); return result; } template<> inline std::string boost::lexical_cast(std::wstring source) { std::string result(source.size(), char(0)); typedef std::ctype<wchar_t> ctype_t; const ctype_t& ct = std::use_facet<ctype_t>(std::locale()); ct.narrow(source.data(), source.data() + source.size(), '\u00B6', &(*result.begin())); return result; }

2 1

Re: [Boost-users] minimal spanning trees & Visual C++ 7.1
by Jasper, Alexander 25 Jul '05

25 Jul '05

Hi! I suffered from ICEs using boost as well. They occured using multi_index lib, though. Turning of "Minimal Rebuild" helped for me. kind regards Alex > -----Ursprüngliche Nachricht----- > Von: boost-users-bounces(a)lists.boost.org > [mailto:boost-users-bounces@lists.boost.org]Im Auftrag von > Amit K. Jain > Gesendet: Montag, 25. Juli 2005 00:55 > An: boost-users(a)lists.boost.org > Betreff: [Boost-users] minimal spanning trees & Visual C++ 7.1 > > > Hello, I've got the latest version of Boost set up w/ Visual C++ 7.1. > When I try to compile kruskal-example.cpp or prim-example.cpp, I get > an internal compiler error on properties.hpp:216. > > I'm reasonably sure that I have Boost set up properly because I can > use the strongly connected components stuff w/ adjacency lists just > fine. > > Has this happened to anyone else and, if so, does anyone know how to > work around it? Thanks in advance > > -Amit > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users >

1 0

Boost.Regex error: match_results error
by reecky 25 Jul '05

25 Jul '05

hi, when I have built the Boost.Regex libarary,and then I tested the samples of the regex. Then I found the credit_card_example correct while the capture_example error. Fistly, the capture_example compiled with the error the .captures(i) is not the member function of the match_results, then I found the source file of match_results.hpp and found this MACRO as below: #ifdef BOOST_REGEX_MATCH_EXTRA typedef typename sub_match<BidiIterator>::capture_sequence_type capture_sequence_type; const capture_sequence_type& captures(int i)const { return (*this)[i].captures(); } #endif so I defined the macro BOOST_REGEX_MATCH_EXTRA, then it compiled successful. but the error occured when I runned the program. the memory leak and system dead. so I wondered what's mean of the BOOST_REGEX_MATCH_EXTRA? many thanks for you. leecky 2005.7.25

2 1

Troubling building Boost for Borland C++ Builder 6
by Edward Diener 25 Jul '05

25 Jul '05

When building Boost for C++ Builder 6 and Boost 1.32, the serialization libray fails to build. The reason for this is that it attempts to use the latest Spirit header files and this does not support C++ Builder 6. What makes this doubly annoying is that the only reason that serialization needs to use Spirit is to support XML serialization. So even if I will never be using serialization for XML serialization, but will use it for the more mainstream binary serialization, I can not use this library with C++ Builder 6. Clearly this is a case for a built library to be split up into sub-libraries, so that there would be a built serialization library supporting binary serialization, which would not be dependent on Spirit, and a built serialization library supporting XML serialization, which would be dependent on Spirit. In this case I would be able to build the library that did not depend on Spirit for C++ Builder and I would not care, if I never used XML serialization, whether or not the library that was built to support XML serialization could be built for C++ Builder 6. In the serialization documentation it says that one can build serialization so that it can be used with C++ Builder 6 by downloading the standalone version of Spirit 1.6, installing that, and setting the SPIRIT_ROOT environment variable to the directory where Spirit 1.6 is installed. Despite having done this, any attempt to build serialization for C++ Builder 6 ends with error messages showing that header files for Spirit in the main Boost directory, rather in the Spirit 1.6 directory, are being used in the process to build the serialization library. The error message comes from spirit/core/config.hpp in my Boost 1.32 directory tree that C++ Builder 6 is not supported. In the Spirit 1.6 distribution there is no spirit/core/config.h. While I understand that a particular library may not want to support a particular compiler/version because that compiler is deficient in the C++ standard requirments to support that library, I feel that if a library is marked as supporting a compiler/version it should do so. In the case of serialization in Boost 1.32 it is marked as supporting C++ Builder 6 when clearly it does not according to my efforts.

4 12

minimal spanning trees & Visual C++ 7.1
by Amit K. Jain 24 Jul '05

24 Jul '05

Hello, I've got the latest version of Boost set up w/ Visual C++ 7.1. When I try to compile kruskal-example.cpp or prim-example.cpp, I get an internal compiler error on properties.hpp:216. I'm reasonably sure that I have Boost set up properly because I can use the strongly connected components stuff w/ adjacency lists just fine. Has this happened to anyone else and, if so, does anyone know how to work around it? Thanks in advance -Amit

1 0

help with integral type conversion trait
by Slawomir Lisznianski 24 Jul '05

24 Jul '05

Hello, I'm looking for a way to convert an integral type into its unsigned equivalent. For example, assume a trait called to_unsigned_integral such that: to_unsigned_integral<int>::type ---> yields unsigned int to_unsigned_integral<short>::type ---> yields unsigned short to_unsigned_integral<unsigned char>::type ---> yields unsigned char ... Regards, Slawomir

2 2

Missing functions in date_time library
by Sean Rohead 23 Jul '05

23 Jul '05

I am using the boost date_time library and noticed that there is a from_ftime() function to convert a FILETIME into a ptime, but I could not find the reverse operation to_ftime(). Am I missing it somewhere? Likewise, I am using the c_local_adjustor::utc_to_local() function and cannot find the corresponding local_to_utc() function in c_local_adjustor. Thanks, Sean

2 1

Re: [Boost-users] Can't create time_duration with milliseconds onVS.NET 2003
by Sean Rohead 23 Jul '05

23 Jul '05

Thank you! I also discovered that this works: time_duration t(8, 45, 0, milliseconds(100).fractional_seconds()); Sean -----Original Message----- From: boost-users-bounces(a)lists.boost.org [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Jeff Garland Sent: Thursday, July 21, 2005 9:37 PM To: boost-users(a)lists.boost.org Subject: Re: [Boost-users] Can't create time_duration with milliseconds onVS.NET 2003 On Thu, 21 Jul 2005 18:37:32 -0600, Sean Rohead wrote > I am using VS.NET 2003 and attempting to create a time_duration > using hours, minutes, seconds, and milliseconds. > > However, I get an error message regarding the 4th parameter > (milliseconds). > > Here is my code and the error message: > > >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> > // main.cpp > > #include "boost/date_time/posix_time/posix_time.hpp" > > using namespace boost::posix_time; > > void main() > { > time_duration t(8, 45, 00, milliseconds(100)); > } The error is valid. Milliseconds is-a time-duration so you don't pass it as the fourth parameter. Add it instead. Something like: Try this instead: time_duration t(8, 45,0); t += milliseconds(100); or even clearer: time_duration t = hours(8) + minutes(45) + milliseconds(100); Jeff _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

2 1