September 2007 - Boost-users

[boost::serialization] Serializing Dynamically Loaded Libraries - "Unregistered Void Cast"
by Jeshua Bratman 06 Sep '07

06 Sep '07

I already sent a message on this topic a few weeks ago, however, my problem has a changed a bit so I decided to post a new message. I've been trying to get Boost serialization working for dynamically loaded shared libraries. Here's my basic structure simplified: 1. base.hpp - includes an abstract base class called base 2. derived.hpp - includes a derived class from base. includes base.hpp. 3. derived.cpp - includes code for derived class and for dynamically loading it 4. main.cpp - includes base.hpp but NOT derived.hpp. dynamically loads instantiates a base * which points to a derived object from derived.so Now in main I want to be able to serialize my pointer to the loaded derived class: base *obj = create_func(); //obj now points to a derived object obj->test(); //when test is run it correctly outputs the message from derived.cpp const base* const to_serialize = obj; std::ofstream ofs("filename"); boost::archive::text_oarchive oa(ofs); try { oa << (to_serialize); } catch(exception &e) { cout << "Exception: " << e.what() << "\n"; } It compiles fine and I got past the "unregisterd class" problem, but now when I try to serialize the pointer an exception is caught: "Unregistered void cast" This appears to be a problem with registering the derived pointer type from the base type when dynamically loaded from a shared library. However, I do register the derived-base relationship as suggested in the documentation. The serialize method in derived.hpp is as follows: template<class Archive> void serialize(Archive & ar, const unsigned int version) { ar.template register_type(static_cast<derived *>(NULL)); //ar & boost::serialization::base_object<base>(*this); ar & data; boost::serialization::void_cast_register( static_cast<derived *>(NULL), static_cast<base *>(NULL) ); } If anyone knows the cause of this and/or a solution I would be very appreciative. I've been working at this for a long time now and I'm almost ready to write my own simple serialization for my specific purpose. Thank you for your help, Jeshua Bratman -- View this message in context: http://www.nabble.com/-boost%3A%3Aserialization--Serializing-Dynamically-Lo… Sent from the Boost - Users mailing list archive at Nabble.com.

4 7

[serialization, ASIO] Assertion when serializing recursive boost::variant type
by Falco Hirschenberger 05 Sep '07

05 Sep '07

Hi, I have defined a recursive variant type like this: --- typedef boost::make_recursive_variant<bool, int, double, std::string, std::vector< boost::recursive_variant_ >, std::map<std::string, boost::recursive_variant_>, TGrey8Image, TRgbImage, TRgbaImage>::type TVType; --- and two containers which can store this type: --- typedef std::vector<TVType> TVector; typedef std::map<std::string, TVType> TMap; --- I wrote a set of unit-tests testing all types of plain and recursive containers and they work as expected. I also tested (de-)serialization to a binary archive and there's also no problem. Then I implemented a Message class which contains such a TVType + some other stuff. Serializing and deserializing also works in the tests. Finally I implemented a client and server with ASIO like the ASIO "serialization" example. First sending the size of the archive and then the data itself. The sending works but when I want to deserialize my message I get the following assertion: --- testSerialization: /usr/include/boost/variant/detail/visitation_impl.hpp:203: typename Visitor::result_type boost::detail::variant::visitation_impl(int, int, Visitor&, VPCV, mpl_::true_, NBF, W*, S*) [with W = mpl_::int_<20>, S = boost::detail::variant::visitation_impl_step<boost::mpl::l_iter<boost::mpl::l_end>, boost::mpl::l_iter<boost::mpl::l_end> >, Visitor = boost::detail::variant::invoke_visitor<boost::detail::variant::direct_assigner<const int> >, VPCV = void*, NBF = boost::variant<boost::detail::variant::recursive_flag<bool>, int, double, std::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::vector<boost::recursive_variant_, std::allocator<boost::recursive_variant_> >, std::map<std::basic_string<char, std::char_traits<char>, std::allocator<char> >, boost::recursive_variant_, std::less<std::basic_string<char, std::char_traits<char>, std::allocator<char> > >, std::allocator<std::pair<const std::basic_string<char, std::char_traits<char>, std::allocator<char> >, boost::recursive_variant_> > >, CIS::Image2d<unsigned char>, CIS::Image2d<CIS::RGB>, CIS::Image2d<CIS::RGBA>, boost::detail::variant::void_, boost::detail::variant::void_, boost::detail::variant::void_, boost::detail::variant::void_, boost::detail::variant::void_, boost::detail::variant::void_, boost::detail::variant::void_, boost::detail::variant::void_, boost::detail::variant::void_, boost::detail::variant::void_, boost::detail::variant::void_>::has_fallback_type_]: Assertion `false' failed. --- I absolutely don't know what't wrong here the backtrace iss huge but the point in my code where it happens is the line when deserializing the message. When I comment out the TVType payload of the message, it works. I would really appreciate any clue. Greetings, Falco

2 1

[Boost.MathToolKit] compilation problem w/ VC7.1 & Boost 1.34.1
by Marc Viala 05 Sep '07

05 Sep '07

Hello, My request is related to the thread http://thread.gmane.org/gmane.comp.lib.boost.user/28535/focus=28613. After downloading & installing Boost 1.34.1 via the Windows Installer, I've tried to compile the following small code snippet: #include <boost/math/special_functions/erf.hpp> void main() { } But the compilation failed w/ the following errors: C:\Program Files\boost\boost_1_34_1\boost\math\special_functions\log1p.hpp(195) : error C2785: 'tools::promote_args<T>::type boost::math::log1p(T)' et 'double boost::math::log1p(double)' ont des types de retour différents C:\Program Files\boost\boost_1_34_1\boost\math\special_functions\math_fwd.hpp(246) : voir la déclaration de 'boost::math::log1p' C:\Program Files\boost\boost_1_34_1\boost\math\special_functions\log1p.hpp(182) : voir la déclaration de 'boost::math::log1p' C:\Program Files\boost\boost_1_34_1\boost\math\special_functions\log1p.hpp(195) : error C2912: spécialisation explicite ; 'double boost::math::log1p<double>(double)' n'est pas une spécialisation d'un modèle de fonction C:\Program Files\boost\boost_1_34_1\boost\math\special_functions\log1p.hpp(183) : fatal error C1903: impossible de récupérer à partir des erreurs précédentes ; arrêt de la compilation Compilation Conditions: - Windows XP Pro - Boost 1.34.1 w/ Math. Toolkit Headers files (extracted from Vault Directory) - Visual C++ 7.1 SP1: Version 7.1.6030 Does anyone know how to manage Math.ToolKit headers file w/ Boost 1.34.1 and VC++ 7.1.6030? Marc Viala, PhD Directeur Général/Directeur Technique 122 rue de la Roche-de-Lorzier 38430 MOIRANS Tél : +33 (0)4 76 91 37 60 Fax : +33 (0)4 76 35 00 07 www.acticm.com mailto:mviala@acticm.com

1 0

How to calculate arboricity
by Abhishek V 05 Sep '07

05 Sep '07

Hi I want to calculate the number of components(arboricity) in a graph. I have created an adjancency list with the help of boost. Is there any function or components avaliable in BGL which can give me the arboricity value. Thanks Abhishek Vyas Tata Consultancy Services Mailto: abhishek.v(a)tcs.com Website: http://www.tcs.com ____________________________________________ Experience certainty. IT Services Business Solutions Outsourcing ____________________________________________ =====-----=====-----===== Notice: The information contained in this e-mail message and/or attachments to it may contain confidential or privileged information. If you are not the intended recipient, any dissemination, use, review, distribution, printing or copying of the information contained in this e-mail message and/or attachments to it are strictly prohibited. If you have received this communication in error, please notify us by reply e-mail or telephone and immediately and permanently delete the message and any attachments. Thank you

2 1

format and filesystem::ofstream mix problem
by Zdravko Baš 05 Sep '07

05 Sep '07

I have a problem with boost 1.34.1 (filesystem::ofstream and format). Compiler Vc71 and vc80 Is this is a bug or my mistake. Zdravko Code: #include <boost/format.hpp> #include <boost/filesystem/fstream.hpp> #include <boost/filesystem.hpp> void test () { boost::filesystem::ofstream os (boost::filesystem::path ("foo.out")); os << boost::format("%s\n") % "foo.text"; } Vc80 output: c:\projects\boost_test\test.cpp(8) : error C2593: 'operator <<' is ambiguous d:\libs\cpp\boost\1.34.1\boost\boost\filesystem\path.hpp(617): could be 'std::basic_ostream<_Elem,_Traits> &boost::filesystem::operator <<<boost::basic_format<Ch>>(std::basic_ostream<_Elem,_Traits> &,const Path &)' [found using argument-dependent lookup] with [ _Elem=char, _Traits=std::char_traits<char>, Ch=char, Path=boost::basic_format<char> ] d:\libs\cpp\boost\1.34.1\boost\boost\format\free_funcs.hpp(33): or 'std::basic_ostream<_Elem,_Traits> &boost::operator <<<char,std::char_traits<char>,std::allocator<_Ty>>(std::basic_ostream<_Elem,_Traits> &,const boost::basic_format<Ch> &)' [found using argument-dependent lookup] with [ _Elem=char, _Traits=std::char_traits<char>, _Ty=char, Ch=char ] while trying to match the argument list '(boost::filesystem::ofstream, boost::basic_format<Ch>)' with [ Ch=char ]

1 0

program_options: change output format?
by Miles Bader 05 Sep '07

05 Sep '07

Is there a way to change the output format that program_options::options_description uses for operator<<? I don't like the output format it uses by default, and I'd like to use the usual GNU format instead. Alternatively, is there a relatively easy way to iterate over all the entries in an options_description (which can be nested, I gather) in a way that I could print the help output myself? From looking over the interface, it seems that there's not really enough information exposed (via the public interface) to do this -- e.g., the "short name" of an option_description is private, and there doesn't seem to be any way to get access to the "groups" in options_descriptions. Thanks, -Miles -- Would you like fries with that?

2 3

Boost.Test claims Qt leaks memory
by Steve M. Robbins 05 Sep '07

05 Sep '07

Hi, I'd appreciate some advice using the Boost unit testing library with Qt classes. I'm using MSVC 8.0 on windows XP. It appears that simply creating and destroying a QObject allocates some static datastructures which are reported as a memory leak by Boost.Test. I'd like to somehow suppress this "static leak" so that Boost.Test may find real leaks in my code. Consider the following test program #define BOOST_AUTO_TEST_MAIN #include <boost/test/auto_unit_test.hpp> #include <Qt/qobject.h> BOOST_AUTO_TEST_CASE( example ) { BOOST_CHECK_EQUAL( 1, 1 ); QObject obj; } It builds and runs fine, except that memory leaks are detected. My first line of attack is to clean up after Qt somehow. (c.f. http://lists.trolltech.com/qt-interest/2007-08/thread00997-0.html) But failing that, is there some way to have Boost.Test ignore these static allocations? I tried turning off leak detection, allocating and deleting a QObject, then turning leak detection back on. But the leaks are still reported. Any ideas? Thanks, -Steve

2 2

tuple versus struct?
by Graham Reitz 05 Sep '07

05 Sep '07

One of my colleagues has asked me, "why not use a simple struct over a tuple?" Unfortunately, I couldn't give a convincing answer, which means I probably don't understand tuple well enough. What is a good way to respond to this? Then we can identify when it makes better sense to use a struct versus a tuple. Thanks, Graham

5 4

using shared_array in Nd array
by Jeffrey Jiang 04 Sep '07

04 Sep '07

Hi, I am new to boost, and is trying to implement shared_ptr solution to a 2d array of pointers. Eg Character *** characters; characters = new Character**[_info._l ]; for (int i=0; i< _info._l; i++){ characters[i] = new Character* [_info._w]; for( int j=0;j < _info._w; j++){ characters[i][j] = 0; } } This is a sparse array, most of the pointers will be set to 0. As I could not find much information about shared_array on line, I would like to check is there any issue with this implementation. boost::shared_array<boost::shared_array<boost::shared_ptr<Character>>> characterGrid(new boost::shared_array<boost::shared_ptr<Character>>[_info._l ]); for( int i=0;i< _info._l; i++) { characterGrid[i] = boost::shared_array<boost::shared_ptr<Character>>(new boost::shared_ptr<Character>[_info._w]); for( int j=0; j< _info._w; j++) { //is it alright for me to use an empty shared_ptr in this context to init the array? characterGrid[i][j] = boost::shared_ptr<Character>(); } } Thanks. Jeffrey.

2 3

Traversing through a multi_array and accessing the data set...
by Graham Reitz 04 Sep '07

04 Sep '07

How do you access the data in the example below? Thanks much, Graham boost::multi_array<double,2> multi_test; boost::multi_array<double,2>::iterator multi_test_iter; boost::multi_array<double,2>::extent_gen extents; multi_test.resize(extents[2][3]); for (multi_test_iter = multi_test.begin(); multi_test_iter != multi_test.end(); ++multi_test_iter) { // How do you get(set) at the current data set pointed to by multi_test_iter? }

2 2