September 2008 - Boost-users

serialization problems (undefined reference - boost::archive::detail::archive_pointer_oserializer<...>)
by Allan Johns 16 Sep '08

16 Sep '08

Hi, I've written my own archive, which works fine until I try serialising a boost::shared_ptr<X>, in which case I get: undefined reference to `boost::archive::detail::archive_pointer_oserializer<boost::archive::boost_oarchive>::~archive_pointer_oserializer(). (boost::archive::boost_oarchive is my archive class, in case this causes confusion). I'm on gcc4.1.2, ceontos4 x86_64. I'm linking to libboost_serialization-gcc41-mt-1_35.so also. I've hit a bit of a brick wall here so any help would be greatly appreciated. I've searched around but can't seem to find any relevant posts on this subject. cheers A _________________________________________________________________ Get all your favourite content with the slick new MSN Toolbar - FREE http://clk.atdmt.com/UKM/go/111354027/direct/01/

3 5

Error while using asio and xpressive together
by Rohit Sauhta 16 Sep '08

16 Sep '08

I get compiler error with the following code: #include <boost/asio.hpp> #include <boost/xpressive/xpressive.hpp> int main(int argc, char*argv[]) {} If I flip the order of #include, code compiles fine. I am using gcc 4.1.2 compiler on linux. Does this need to be fixed in asio ? thx Rohit

2 1

Problem with date_time facets
by Tavares, Pedro (London) 16 Sep '08

16 Sep '08

Hi all, I'm using boost with stlport 5.1.5 in msvc 7.1. when I try to use date facets, the compiler complains there is not suitable std::locale ctor, although I can see it in the code and all flags seem to be defined correctly (_STLP_MEMBER_TEMPLATES is defined). I'm trying to compile the following excerpt: 1 std::stringstream str(__DATE__); // for example "Sep 16 2008" 2 3 boost::gregorian::date_input_facet facet("%b %d %Y"); 4 std::locale loc(str.getloc(),facet); 5 str.imbue(loc); 6 7 boost::gregorian::date d; 8 str >> d; The error is in line 4 and the message is that the (templated) ctor does not exist. Thanks for your assistance, Pedro -------------------------------------------------------- This message w/attachments (message) may be privileged, confidential or proprietary, and if you are not an intended recipient, please notify the sender, do not use or share it and delete it. Unless specifically indicated, this message is not an offer to sell or a solicitation of any investment products or other financial product or service, an official confirmation of any transaction, or an official statement of Merrill Lynch. Subject to applicable law, Merrill Lynch may monitor, review and retain e-communications (EC) traveling through its networks/systems. The laws of the country of each sender/recipient may impact the handling of EC, and EC may be archived, supervised and produced in countries other than the country in which you are located. This message cannot be guaranteed to be secure or error-free. This message is subject to terms available at the following link: http://www.ml.com/e-communications_terms/. By messaging with Merrill Lynch you consent to the foregoing. --------------------------------------------------------

1 0

date-time library and checked iterators
by gast128 16 Sep '08

16 Sep '08

Dear all, last week we switched to vstudio 2008. It has checked iterator support and it already catched two incorrect (but probably also harmless) uses of iterators. Maybe in the date time library there is also such a case. The stream operator on ptime eventually calls date_time::date_facet::do_put_tm () function which probably uses an incorrect way of dereferencing string iterators when empty: // use time_put facet to create final string return std::use_facet<std::time_put<CharT> >(a_ios.getloc()).put(next, a_ios, fill_char, &tm_value, &*a_format.begin(), &*a_format.begin()+a_format.size()); When a_format is empty the begin iterator may not be dereferenced? I do know that the actual code is harmless, but now we get the M$ checked iterator error message, and turning it globally off would be pity, since it cna be an extra potential debugging tool. Wkr, me

1 0

[test] Passing parameters to fixtures
by Markus Schöpflin 16 Sep '08

16 Sep '08

Hello, to set up and tear down database connections for unit tests, I need something like parametrized test fixtures. Up to now (using an old version of Boost.Test) I was creating a customized test suite (by deriving from test_suite) and performed the connection handling during construction/destruction of the customized test suite. As the latest Boost.Test now has explicit support for test fixtures, I was looking at how I could replace my old approach by using a test suite fixture. The fixture would need to look something like this: struct F { F(string db_name) { connect to db_name } ~F() { disconnect from db_name } }; Now, the Boost.Test documentation tells me that a fixture is supposed to be used like this: BOOST_FIXTURE_TEST_SUITE( s, F ) So there doesn't seem to be a way of passing parameters to the constructor of the fixture. Is there any? Or is there any other way to obtain a parametrized fixture? Or should I just keep on using customized test suites? TIA, Markus

4 10

Whatever happened to singleton class?
by Kurt Kohler 15 Sep '08

15 Sep '08

Back in January, the review of Tobias Schwinger's singleton package resulted in it being rejected. At the time it was stated that a new version would be out for review in three months or so. Has anything happened since then? I've been using the class without problems for some time, but I'd feel more comfortable if it was an official part of boost.

2 1

boost.extensions and boost.reflection.
by Germán Diago 15 Sep '08

15 Sep '08

Hello. I'm implementing a property class which uses my own library (still in early stages) and I want to extend it to general reflection (not for now) and shared library loading. I found boost.extensions, which is not a boost library for now. I don't know if this is the place to ask for this, if it is not, point me where I could go. It's great to see I won't have to taje any more pain to load plugins :-). I would like to make a feature request. In boost.extensions, factories are used to construct types of a derived class with pointers. So, in order to register a new class, you have to tell which is the base class to the extensions system. I think there is a better way, with which you wouldn't need to do this. It is inspired in this type style of value-based type system: http://stlab.adobe.com/wiki/index.php/Image:2008_06_26_classes_that_work.pdf An object which can contain any other object (like boost::any) could be created. This way (this is how I do in my own solution) you just have to register a type without worrying about its base class: REGISTER_CLASS(myns, myclass) Now I can do like this: object_t myobject = introspection::create_object("myns", "myclass"); when I want my object to become "myclass" I do: myns::myclass & refclass = get_reference<myns::myclass>(myobject); This works fairly well, and factories go away, freeing the user from learning more interfaces. Just register your type and use it. Everything else is transparent. The interface is simple, no factories or anything else to handle. This can be also used from shared libraries without problems. At least, I think that it should be provided a way to register a factory of non-pointer types. In the documentation it says a pointer is created for the animal class, disabling this type of value-based registration.

2 4

fusion::result_of::transform<Seq1,Seq2,F>
by e r 15 Sep '08

15 Sep '08

hello, here's what I'm trying to do: *.cpp //result_type() not valid typedef fusion::vector<A,B,C> result_type; //arg2s_type() is valid typedef fusion::vector<Arg2A,Arg2B,Arg2C> arg2s_type; //transforms // (fusion::vector<identity<A>,identity<B>,identity<C> >(),arg2s_type()) // into {A(arg1,Arg2A()),B(arg1,Arg2B()),C(arg1,Arg2C()))} result_type result( range<result_type,args2_type>::create( arg1 //arbitrary type Arg1 ) ); here's my attempt. problems/questions are shown in comments. *.hpp: template<template<typename> class metafun = mpl::identity> struct apply_wrapper{ template<typename T> struct apply{ typedef metafun<T> type; }; }; template<typename Seq,template<typename> class metafun = mpl::identity> struct apply_wrapper_to_seq{ typedef typename mpl::transform< Seq, // e.g. fusion::vector<A,B,C> apply_wrapper<metafun> >::type type; //e.g. fusion::vector<identity<A>,identity<B>,identity<C> > }; template<typename Arg1> struct fun{ fun(const Arg1& args):args_(args_){} //copy, assign... // probably not what result_of::transform expects, based on compile // errors (below) template<typename Metafun,typename Arg2> struct result{ typedef typename Metafun::type type; }; template<typename Metafun,typename Arg2> typename Metafun::type //e.g Metafun = identity<A> operator()(const Metafun& meta,const Arg2& arg2)const{ typedef typename Metafun::type result_type; //e.g. A return result_type(args,arg2); } const Arg1& args_; }; template<typename Result,typename Arg2s> struct range{ //e.g. fusion<identity<A>,identity<B>,identity<C> > typedef typename apply_wrapper_to_seq<Result>::type wrappers_type; //e.g. fusion::vector<Arg2A,Arg2B,Arg2C> typedef Arg2s arg2s_type; template<typename Arg1> struct result{ typedef typename fusion::result_of::transform< wrappers_type, arg2s_type, fun<Arg1> >::type type; }; template<typename Arg1> static typename result<Arg1>::type create(const Arg1& args){ typedef fun<Arg1> f_type; wrappers_type wrappers; arg2s_type arg2s; f_type f(args); BOOST_MPL_ASSERT( ( mpl::equal< Result, typename result<Arg1>::type > ) );//result_of::fun causes compile error /usr/local/boost_1_35_0/boost/utility/result_of.hpp|34|instantiated from ‘boost::detail::result_of_impl< factory::fun<Arg1>, factory::fun<Arg1> > ()(boost::mpl::identity<A>, Arg2A), false>’| ... //usr/local/boost_1_35_0/boost/utility/result_of.hpp|68|error: wrong number of template arguments (1, should be 2)| //factory/range.hpp|32|error: provided for ‘template<class Arg1> template<class Metafun, class Arg2> struct factory::fun<Arg1>::result’| //usr/local/boost_1_35_0/boost/fusion/view/transform_view/detail/value_of_impl.hpp|55|error: no type named ‘type’ in ‘struct //boost::mpl::apply<boost::fusion::detail::apply_transform_result<factory::fun<Arg1>, boost::mpl::identity<A>, Arg2A, mpl_::na, mpl_::na, mpl_::na>’| return fusion::transform( wrappers, arg2s, f ); } }; any help would appreciated. thanks!

2 7

boost.extensions bugs?
by Germán Diago 15 Sep '08

15 Sep '08

Hello. I downloaded boost.extensions from http://boost-extension.blogspot.com/?gclid=CNjp5dfJ2JUCFQsZQgod0mHbWg When I tried the first example (the shared_library one) in the tutorial, I realized that the macro BOOST_EXTENSION_EXPORT_DECL didn't prevent c++ name mangling for the example function. I had to add extern "C" before void manually. And in the second tutorial (the animals tutorial) BOOST_EXTENSION_TYPE_MAP_FUNCTION seems to be called, but when I get the map from main, it is empty. The library is found (LD_LIBRARY_PATH is set so that it finds the library) but it does not work. Thank you. Tested in gcc 4.2.3 under ubuntu 8.04.

2 3

Memory/CPU usage
by Jean-Sebastien Stoezel 15 Sep '08

15 Sep '08

Hi, Is there a library in Boost that provides memory/CPU usage estimates? Thanks, Jean

2 1