February 2006 - Boost-users - lists.preview.boost.org

[boost][serialization] How to implement serialization classes in two DLLs
by Marc Viala 02 Feb '06

02 Feb '06

Hi, I'm very sorry to post the message hereafter a second time, but I've not received any reply, maybe I've missed it. Any workaround or recommendations will be welcome because we are right now surrounded in our software development. Thanks in advance. Marc VIALA After few hour on the subject, I'm would like to have some explanations & recommendations one the boost serialization library. To explain, my problem, you will find hereafter a VC++ 7.1 project that demonstrates it. Problem Description We have involved in a large project with for example two DLLs libraries: LibA and LibB. These libraries define & implement one class, let's say libA::A and libB::B. These two classes can be serialized through boost serialization engine. The interfaces for the two classes are the same: ---------------------- struct LIBA_API A { std::vector<double> my_vector ; private: friend class boost::serialization::access; template<class Archive> void save(Archive & ar, const unsigned int version) const ; template<class Archive> void load(Archive & ar, const unsigned int version) ; BOOST_SERIALIZATION_SPLIT_MEMBER() }; LIBA_API void save(const libA::A* pObj, std::ostream& os) ; LIBA_API void load(libA::A* &pObj, std::istream& is) ; }; // Namespace libA ---------------------- The two free functions save() and load() are defined to be able to serialize the class libA::A (respectively the class libB::B). When, we try to use these interfaces in an EXE application: ---------------------- #include <LibA/A.h> #include <LibB/B.h> #include <boost/filesystem/path.hpp> #include <boost/filesystem/fstream.hpp> #pragma comment(lib, "liba_d.lib") #pragma comment(lib, "libb_d.lib") int main(int argc, char* argv[]) { libA::A a; libA::save(&a, boost::filesystem::ofstream("a.xml")); libB::B b; libB::save(&b, boost::filesystem::ofstream("b.xml")); // -> Assert in debug mode return 0; } ---------------------- The figure hereafter describes the relationships between LibA, LibB and EXE application: ------------ ------------ | LibA | | LibB | ------------ ------------ \ / \ / Exe Application We have to face to an assert in debug mode generate by the file boost::serialization::extended_type_info.cpp, line#71. This assert happens during the instance serialization of class libB:B. Recently, there are some posts on this subject, but I would like to known how I can deal with this assert? Is it possible to avoid it by a code organisation? Is it possible to comment this assert? To have a concrete answer, I have joined my complete VC++7.1 project. Thanks in advance for your help. Development Platform WindowsXP VC++ 7.1 Boost 1.33.1 Marc Viala mailto:mviala@acticm.com <mailto:mviala@acticm.com>

3 2

Re: [Boost-users] trying to install Boost 1.33.1 on Solaris 5.9 + SunStudio 11
by paul.floyd 02 Feb '06

02 Feb '06

... > If found this in the sun forum. > http://blogs.sun.com/roller/page/sga > Contains a patch to boost 1.33.1 and 1.32.0 to help compiling on sun. Very interesting. Can anyone comment as to whether these changes are going to be integrated to boost? They haven't been yet (at least in the code that I can browse with the web interface). A+ Paul Accédez au courrier électronique de La Poste : www.laposte.net ; 3615 LAPOSTENET (0,34 /mn) ; tél : 08 92 68 13 50 (0,34/mn)

3 5

[program_options] Uncatchable Errors
by Merrill Cornish 02 Feb '06

02 Feb '06

I have a minimal program below that uses the program_options library. If executed with no options or with --help as an option, it correctly prints "End of option processing" and does nothing else. However, if you give it an unexpected option such as --x, then (in my case) Zone Alert pops up saying that "somone" is trying to launch Dr Watson. If the try/catch is omitted, a message is printed to the console: "This application has requested the Runtime to terminate it in an unusual way. Please contact the application's support team for more information." Am I doing something wrong to make unexpected options behave so badly? Merrill ========================================== #include <iostream> #include <stdexcept> #include "boost/program_options.hpp" namespace po = boost::program_options; using namespace std; int main(int argc, char* argv[]) { try { po::options_description general_verbose("Options"); general_verbose.add_options() ("help,h", " Display this help message."); po::variables_map vm; po::store(po::command_line_parser(argc, argv). options(general_verbose).run(), vm); po::notify(vm); }//try catch(std::exception ex) { cout << "Exception: " << ex.what() << endl; } catch(...) { cout << "Unknown exception." << endl; } cout << "End of option processing." << endl; }//main()

1 0

[python] create a custom exception
by boost.users＠cgrealy.mailcan.com 02 Feb '06

02 Feb '06

Is it possible to create a C++ to python exception translator that raises a custom python exception? I know you can use register_exception_translator to translate from a c++ exception to a standard python exception, but is it possible to raise a user-defined exception? Regards, Colin

2 1

Re: [Boost-users] [serialization] compile error
by cheesy4poofs 02 Feb '06

02 Feb '06

> I compiled the following program wiht no errors on my vc 7.1 > system with no errors or warnings > > Robert Ramey I finally solved my problem by commenting out the following include: //#include <boost/serialization/base_object.hpp> #include <boost/archive/text_iarchive.hpp> #include <boost/archive/text_oarchive.hpp> #include <boost/serialization/map.hpp> Not sure why it solved it, but I'm glad at least to have put this behind me. Thanks for the help. Scott

3 2

[serialization] Removing empty classes
by Loïc Joly 01 Feb '06

01 Feb '06

Hello, Let's say I have a structure suce as this : class A {}; class B : public A {}; class C : public B {}; Currently, I have some code that looks like this : template<class Archive> void C::serialize(Archive &ar, const unsigned int version) { ar & BOOST_SERIALIZATION_BASE_OBJECT_NVP(B); } However, if my class B is "empty", that is it adds no data to A, is this possible as an optimisation to just write : template<class Archive> void C::serialize(Archive &ar, const unsigned int version) { ar & BOOST_SERIALIZATION_BASE_OBJECT_NVP(A); } Or will this break something ? -- Loïc

2 1

Re: [Boost-users] [serialization] compile error
by cheesy4poofs 01 Feb '06

01 Feb '06

>Try removing the line > >> #include <boost/serialization/type_info_implementation.hpp> > >And let me know that happens. > >I believe this problem is a remnant of a header depency issue which will be addressed in > the next release. > >Robert Ramey Thanks for the response. I actually didn't have that #include in my original code. I added it in a futile attempt to see if that would fix the issue. I still see the same compile errors when I remove that #include line. Scott

2 1

[Serialization]: bug while serializing a std:string
by RIVASSEAU Jean Noel 01 Feb '06

01 Feb '06

Hello It seems I have come against a bug in the BSL. I am running on Win 2000 Pro, MinGW (gcc 3.4.2), Boost 1.33.1. If I serialize a standard string with 512 or more characters, the save goes ok but a stream error exception occurs while trying to reload it later from archive. Any string with size < 512 is ok. Also this bug is really strange. It does not happen if the string has no line breaks in it (even if it is more than 512 chars). It also happens on my program, but I was unable to reproduce it with a small test program, even using the same string. Could it be a bug in another part of my rather large program? Or is it really a bug in the BSL? Jean-Noël

2 1

Bug with std::string in the BSL
by RIVASSEAU Jean Noel 01 Feb '06

01 Feb '06

Hello Following my previous message, I have not been able to track down the bug. What I noticed is that the stream error occurs after the loading of the std::string (on the next load operation), however the loaded string is not correct (eg, even if the exception is thrown later, the problem lies with the strings of length >= 512). Robert, if you cannot look into that, can you give me some pointers about where should I look to debug this problem? I am using a text archive. Hints about the implementation code would be useful for me. Thank you Jean-Noël

1 0

[date_time] Constructing a ptime with __int64
by Christian Henning 01 Feb '06

01 Feb '06

Hi there, is it possible to construct a ptime with an __int64. The internal representation is of the same type. I couldn't find anything on this issue. Thanks in advance, Christian

2 4