May 2009 - Boost-users - lists.boost.org

[Boost.serialization] unexpected bad_cast exception while deserializing some arbitrary instances (e.g.: ar & data) ...
by Francois Mauger 28 May '09

28 May '09

Hi serializers! Let me present my problem: I have a class, named 'io_factory', that enables to deserialize an arbitrary object from a text (or XML) archive using the following '__load_text' template method: -8<--------------------------------------------------------------------- class io_factory { ... template <typename Data> void __load_text (boost::archive::text_iarchive & ar_, Data & data_) { Data & b = data_; std::cerr << "DEVEL: io_factory::__load_text: archive class name='" << typeid (ar_).name () << "'" << std::endl; std::cerr << "DEVEL: io_factory::__load_text: class name='" << typeid (b).name () << "'" << std::endl; try { /* Here I may experience an unexpected exception * only in very special circumstances: */ ar_ >> b; std::cerr << "DEVEL: io_factory::__load_text: deserialization is ok." << std::endl; } catch (std::exception & x) { std::cerr << "DEVEL: io_factory::__load_text: EXCEPTION=" << x.what () << std::endl; } std::cerr << "DEVEL: io_factory::__load_text: done" << std::endl; } ... }; -8<--------------------------------------------------------------------- I wrote a shared library that makes use of this template code segment. Using a test program, as I try to deserialize an instance of some 'nemo3::ana_event' class from a Boost text archive file, it prints: -8<--------------------------------------------------------------------- DEVEL: io_factory::__load_text: archive class name='N5boost7archive13text_iarchi veE' DEVEL: io_factory::__load_text: class name='N5nemo39ana_eventE' <= ok 'nemo3::ana_event' DEVEL: io_factory::__load_text: deserialization is ok. <==== IT WORKS! DEVEL: io_factory::__load_text: done -8<--------------------------------------------------------------------- This is fine! My object is properly loaded from the file. My test program works as expected. Well, now, I also have a Python module that wraps some C++ code and the 'io_factory' class. I use Boost.Python to do that. When I run my Python program, importing this wrapper module, to deserialize the same Boost text archive file from the python shell; it prints: -8<--------------------------------------------------------------------- ... DEVEL: io_factory::__load_text: archive class name='N5boost7archive13text_iarchiveE' DEVEL: io_factory::__load_text: class name='N5nemo39ana_eventE' DEVEL: io_factory::__load_text: EXCEPTION=std::bad_cast <= OOOOPS! DEVEL: io_factory::__load_text: done ... -8<--------------------------------------------------------------------- As you can see, there is a nasty 'std::bad_cast' thrown exception from the "ar_ >> b;" statement. This code works perfectly in the first case, now something is broken and I cannot figure out what! I am really confused by this issue. I already used exactly the same approach within another library (also based on Boost.serialization and Boost.Python) and I never met any problem. More, I also have a smaller python script that uses the same wrapper module with the same input archive file and it works: I can read all the instances I stored in my archive file! If I use an XML archive, the problem is exactly the same: it works in the first case, not in the last one. I do not understand where this 'bad_cast' exception comes from. For me it as nothing to do with: - the format of the archive (I know the archive file is ok as the test program store/load is properly) - the wrapping within Python (I can make it run ok within a sample python script) I really need a hint that could help me to understand this issue. It seems this problem looks like a side-effect (?) anyway the deep reason is outside of the scope of my code and my knowledge. My config is: * OS: Ubuntu Linux 8.04 * Linux Kernel: $ uname -a Linux mauger-laptop 2.6.24-24-generic #1 SMP Wed Apr 15 15:54:25 UTC 2009 i686 GNU/Linux * Boost: $ boost-config --version 1_38_0 * gcc: $ g++ --version g++ (GCC) 4.2.4 (Ubuntu 4.2.4-1ubuntu3) * Python 2.5 I'd really appreciate your comments. I'm afraid I cannot send a small sample code that could reproduce this 'bug' in a simpler way as it comes only at a quite high level on top of 3 big shared libraries + Boost + Python. My attempt to isolate some critical code to see this side-effect within "Boost.serialization only" or "Boost.Python only" failed. More it confirms that the code works as expected (hum... or seems to work...). Thanks a lot for your help. regards frc -- François Mauger Département de Physique - Université de Caen Basse-Normandie courriel/e-mail: mauger(a)lpccaen.in2p3.fr tél./phone: 02 31 45 25 12 / (+33) 2 31 45 25 12 fax: 02 31 45 25 49 / (+33) 2 31 45 25 49 Adresse/address: Laboratoire de Physique Corpusculaire de Caen (UMR 6534) ENSICAEN 6, Boulevard du Marechal Juin 14050 CAEN Cedex FRANCE

4 5

[boost::serialization] string serialization
by Bob Walters 28 May '09

28 May '09

I've been trying to use boost::serialization to serialize non-standard strings. Specifically, std::basic_string with a non-standard allocator type, and in another case, boost::interprocess::basic_string. Looking at how std::string serialization was enabled in serialization/string.hpp, I thought I would try to broaden that approach by using a specialization of the implementation_level template as follows: namespace boost { namespace serialization { template <typename C, class T, class A> struct implementation_level< std::basic_string<C,T,A> > { typedef mpl::integral_c_tag tag; typedef mpl::int_< boost::serialization::primitive_type > type; BOOST_STATIC_CONSTANT( int, value = implementation_level::type::value ); }; } // namespace } // namespace I've tried also adding wraper traits (also in boost::serialization namespace): template<typename C, class T, class A> struct is_wrapper< std::basic_string<C,T,A> > : mpl::true_ {}; Serialization of these strings is compiling, and running. At first it looked like it worked, but not quite. Specifically: when the string in question is 16 bytes or less in size (including the \0), the string serializes fine. When longer than that, its value appears to be completely truncated. (I'm looking at the dump of a binary_oarchive.) This isn't the case for std::string, but I'm having a problem figuring out why. I realize this is an unconventional scenario, but any assistance on how to make the support for string serialization apply to more than just std::string is appreciated. (Perhaps this should be ticketed as a feature request?)

1 0

[ASIO] Deadlock during static destruction
by Dan Eaton 28 May '09

28 May '09

Hello, I'm using boost::asio on WinXP64, with boost 1.39. My app is deadlocking during destruction of a static object that has an io_service member variable. If I explicitly delete the static object, no deadlock occurs. The static object is actually a shared_pointer, which points to the object I mentioned above. So at exit, assuming I haven't explicitly deleted the object, once the reference count reaches zero, delete is called, and the process deadlocks. The offending piece of code appears to be in win_thread.hpp:143. void join() { ::WaitForSingleObject(exit_event_, INFINITE); ::CloseHandle(exit_event_); if (terminate_threads()) { ::TerminateThread(thread_, 0); } else { ::QueueUserAPC(apc_function, thread_, 0); ::WaitForSingleObject(thread_, INFINITE); } } It is blocked at ::WaitForSingleObject(exit_event_, INFINITE); (This is as a result of the destruction of the io_service member). In case it's significant, the io_service object is used to establish several socket connections (to and from remote hosts), some of which are through socket.connect, and others through socket.connect_async. Thanks! Danny

2 1

[thread] getting thread id number
by Hipatia 27 May '09

27 May '09

Hi! I'm trying to get the id number of a thread just to write it to the standard output. Since I'm not an expert in c++ programming, let me explain what I do (even if it's quite simple): a create an instance as follows, boost::thread::id thread_id= boost::this_thread::get_id(); By debugging my code with Visual Studio, I see that 'thread_id' has an 'unsigned int' object member called 'id', so I would like to display this value. Unfortunately, I cannot find in the boost documentation any function member of boost::thread::id class that returns its value. Thanks, Hipatia. -- View this message in context: http://www.nabble.com/-thread--getting-thread-id-number-tp23222717p23222717… Sent from the Boost - Users mailing list archive at Nabble.com.

5 6

[threads] calling unlock twice
by Ovanes Markarian 27 May '09

27 May '09

Hello *, I wrote a code which used a RAII idiom to lock multiple mutex instances in ctor and unlock them in dtor. This resulted in the class similar to: struct lock_many { lock_many(mutex& m1, mutex& m2, ...) : m1_(m1), ... { boost::lock(*m1_, *m2_, ...); } ~lock_many() { //some asserts here, which ensure that either all mutex pointers are NULL or unequl to each other if(m1_==NULL && m2_==NULL && ...) return; // all are NULL no unlock needed } private: mutex* m1_; mutex* m2_; };

3 3

[asio] socket_types.hpp defines _WIN32_WINNT under Windows Mobile
by Tom Aylesworth 27 May '09

27 May '09

The socket_types.hpp defines _WIN32_WINNT to 0x0501 if neither _WIN32_WINNT nor _WIN32_WINDOWS are defined. However, under Windows Mobile, _WIN32_WINNT should not be defined. It appears this is only being used to determine the value for max_iov_len: # if defined(_WIN32_WINNT) const int max_iov_len = 64; # else const int max_iov_len = 16; # endif So I think that the original test should also test for _WIN32_WCE, so that _WIN32_WINNT is not defined under Windows Mobile: # if !defined(_WIN32_WINNT) && !defined(_WIN32_WINDOWS) && !defined(_WIN32_WCE) Does anyone know what max_iov_len should be set to under Windows Mobile 5, or how I might be able to find out? I assume it's safer to leave it at the lower value (16) for now, but it would be nice to have it set correctly. Thanks, Thomas Aylesworth Senior Software Engineer RealityMobile http://www.realitymobile.com/ 703.636.7202 office 703.209.1081 mobile

1 0

[Boost.MPI] SendingReceiving an array "view"
by Joel Falcou 27 May '09

27 May '09

I have a small class view<T> that containes a T* and a size. When I send/receive view using Boost.MPI, I want transfert the actual n data pointed by the view pointer and send the size. On receiving, i will receive the size, checks if they are compatible and receive the data into the zone pointed by the pointer of the receiving view. Of course, I may do something more complex than send/Receive, like all_2_all etc. So i was thinking to give view<T> a proper serialize function so boost.serialization may kick in. As the behavior of sending and receiving is different, I use the load/save idiom to separate the behavior. The question is then, how to serialize the n elements poitned by a raw pointer to those data in an efficient way ? Regards

4 4

Integral type with no upper limit?
by Robert Dailey 27 May '09

27 May '09

Is there an integral type in Boost that has no upper limit? The only physical limit would be the amount of memory on the system. And of course, the larger the number, the heavier the math is on such an object, because more cycles are required.

4 4

transform_iterator causing strange problems
by Lindley M French 26 May '09

26 May '09

I'm not sure if this is the fault of transform_iterator----I really don't see how it could be----but I'm really confused here. I've got an unordered_map<GroupID, Group> (typedefed to GroupMap) in my class. The specifics of those types don't matter except that GroupID is an integral type, and Group contains a std::vector vec. I'm trying to use a transform_iterator so that my class may be viewed as a group container: typedef boost::transform_iterator<GroupSelector, GroupMap::const_iterator> const_iterator; const_iterator begin() const { return boost::make_transform_iterator(currgroups.begin(),GroupSelector()); } const_iterator end() const { return boost::make_transform_iterator(currgroups.end(),GroupSelector()); } std::pair<const_iterator, const_iterator> groups() const { const Group &grp = *(begin()); return std::make_pair(begin(),end()); } Where my GroupSelector class is defined as struct GroupSelector { typedef const Group& result_type; const Group& operator()(const std::pair<GroupID, Group> &p) const { std::cout << p.second.vec.size() << " " << &p.second << std::endl; return p.second; } }; Note the output statement. Now, the usage in my main function is for (tie(iter,end) = final.groups(); iter != end; ++iter) { const Group& grp = *iter; cerr << &grp << endl; cerr << grp.segs.size() << endl; ... My output is 10 0012F50C 10 0012F65C 0012F65C 1243176 which is strange for two reasons. The first line comes from GroupSelector when I dereference begin() in the groups() function. The second comes from GroupSelector when I dereference iter in main. As these should both correspond to the same iterator, I can't understand why the address of the returned Group reference is different. The 3rd and 4th lines are outputted in main. Here's the second confusion: The address for the Group object is the same as in the most recent GroupSelector output, yet now the size of the vector within the Group comes out wrong. HUH?

5 8

Re: [Boost-users] [BGL] Trying to use subgraph
by Lindley M French 26 May '09

26 May '09

My inbox was full so I had to grab the response off the archive, hence the lack of quote. However, this: typedef boost::property<edge_index_t, size_t, EdgeProperties> EdgePropWithIndex; .... typedef boost::subgraph< boost::adjacency_list< boost::vecS,// Use list for edges boost::vecS,// Use vector for nodes boost::undirectedS,// Undirected graph SegIndex,// Nodes are segments EdgePropWithIndex// Edges have weights > > BoostGraph; still gives me compile errors: error C2065: 'edge_index_t' : undeclared identifier error C3203: 'property' : unspecialized class template can't be used as a template argument for template parameter 'EdgeProperty', expected a real type On a (possibly) related note, I'm trying to use the above subgraph in kruskal_minimum_spanning_tree(), but it's giving me an error about error C2039: 'vertex_bundled' : is not a member of 'boost::subgraph<Graph>' I'm quite sure this error didn't appear before I started trying to use the subgraph adapter.....any idea what's wrong?

2 19