March 2007 - Boost-users - lists.boost.org

Serialization: Serialize into multiple archives
by Stefan Holzmann 23 Mar '07

23 Mar '07

Hello, I'am new to the boost serialize libary and have the following question. At my data acquisition application I have a workspace object which contain several recorded data objects (measurements) which should be stored into separate archives. The workspace objects needs to store references (weak pointers) on objects inside the measurement archives (e.g. references from views which should be serialized). It should be possible to have the same measurement archive references from more than one workspace object. Is this possible with the boost serialization libary? Righ now I don't know how the object tracking is working in detail, is there a more detailed documentation on this? Stefan

1 0

Re: [Boost-users] Question about ifstream ranges
by Pfligersdorffer, Christian 23 Mar '07

23 Mar '07

boost-users-bounces(a)lists.boost.org wrote: > Pfligersdorffer, Christian wrote: >> Hello all! >> >> I face the problem of partitioning a file that i read with > ifstream into portions that are to be processed by some > clients. I want to ensure that a client only views a specific > part (window, range) of the stream. So far I store 2 > streamoffsets posStart and posEnd, call seekg on the stream, > pass it to the client by ref and rely on the client to check > when it reached posEnd. >> >> However, I certainly cannot copy the whole range into a new > stream object and pass that one. Efficiency does matter. >> >> Any suggestions? Maybe some iterator-magic can be done? > > How about Boost.Spirit file_iterator? > That is a read-only RandomAccessIterator for files. > http://spirit.sourceforge.net/distrib/spirit_1_8_3/libs/spirit > /doc/file_iterator.html Hello Shunsuke, that's a good idea but unfortunately I cannot use this class in my case. I have some special decoding ifstreams doing decompression and the like. So I'd really need an iterator that takes an i(f)stream object in it's constructor and wraps that. Thanks anyway! -- Christian Pfligersdorffer Dipl.-Ing. Bakk.rer.nat. Software Engineering EOS GmbH - Electro Optical Systems Robert-Stirling-Ring 1 D-82152 Krailling / Munich Phone +49 89 893 36-229 Fax +49 89 893 36-287 christian.pfligersdorffer(a)eos.info http://www.eos.info

4 3

[asio] any protocol libs built on top?
by Sohail Somani 23 Mar '07

23 Mar '07

Hi, Have there been any implementations of some common protocols using asio? I know its still early days, but just wondering. Thanks, Sohail

2 2

[asio] Compilation requirements
by Sohail Somani 23 Mar '07

23 Mar '07

Hi, Is asio is a header-only library? If so, what are the requirements for compilation? If you are using Visual C++ with /Za (language extensions disabled), that will cause problems if you include windows.h. TIA, Sohail

4 7

[shared_ptr] why segfault?
by Xu, Peng 23 Mar '07

23 Mar '07

greetings, my test code is as follows: " int k = 128; std::vector<boost::shared_ptr<int> > v(k); for(unsigned int i = 0; i < k; ++i) { v[i].reset(&k); // v[i].reset(new int(0)); } int t = rand() % k + 1; std::cout << "element " << t << " points to integer value " << *(v[t].get()) << std::endl; } " It's intended to use shared_ptr in an atypical way(the normal one was commented out). The output is: " element 68 points to integer value 128 Segmentation fault " Attached debugger and got this stack: " Program received signal SIGSEGV, Segmentation fault. 0x0000002a95c16ed0 in _int_free () from /lib64/tls/libc.so.6 (gdb) bt #0 0x0000002a95c16ed0 in _int_free () from /lib64/tls/libc.so.6 #1 0x0000002a95c17317 in free () from /lib64/tls/libc.so.6 #2 0x0000002a95862cee in operator delete () from /usr/lib64/libstdc++.so.5 #3 0x0000000000403ac7 in boost::checked_delete<int> (x=0x7fbfffe9c8) at checked_delete.hpp:34 #4 0x00000000004035b4 in boost::detail::sp_counted_impl_p<int>::dispose (this=0x509060) at sp_counted_impl.hpp:78 #5 0x0000000000402cf8 in boost::detail::sp_counted_base::release (this=0x509060) at sp_counted_base_gcc_x86.hpp:146 #6 0x0000000000402dee in _ZN5boost6detail12shared_countD9Ev (this=0x508058) at shared_count.hpp:199 #7 0x0000000000402e0d in boost::detail::shared_count::~shared_count (this=0x508058) at shared_count.hpp:65536 #8 0x0000000000402831 in destr_detour25 () at shared_ptr.hpp:65536 #9 0x0000000000402855 in boost::shared_ptr<int>::~shared_ptr (this=0x508050) at sp_counted_base_gcc_x86.hpp:110 #10 0x0000000000403822 in std::_Destroy<boost::shared_ptr<int> > (__pointer=0x508050) at stl_construct.h:125 #11 0x00000000004037f6 in std::__destroy_aux<boost::shared_ptr<int>*> (__first=0x508050, __last=0x509050) at stl_construct.h:101 #12 0x00000000004037c7 in std::_Destroy<boost::shared_ptr<int>*> (__first=0x508050, __last=0x509050) at stl_construct.h:143 #13 0x0000000000401d3f in _ZNSt6vectorIN5boost10shared_ptrIiEESaIS2_EED9Ev (this=0x7fbfffe900) at stl_vector.h:297 #14 0x0000000000401da1 in std::vector<boost::shared_ptr<int>, std::allocator<boost::shared_ptr<int> > >::~vector (this=0x7fbfffe900) at stl_vector.h:65536 #15 0x000000000040160a in destr_detour44 () at shared_ptr.cpp:42 #16 0x0000002a95bcaacd in __libc_start_main () from /lib64/tls/libc.so.6 #17 0x000000000040108a in _start () at start.S:96 " I expected some problems with the code, but the segfault surprised me. Can anybody tell me why? Thanks a lot,

5 7

[mpl][fusion] Metafunction for extracting a type from a map
by Christian Henning 23 Mar '07

23 Mar '07

Hi there, next metaprogramming problem. Is it possible to extract a tagged type from a map? namespace fields { struct _id_; struct _name_; struct _birthday_; struct _address_id_; } namespace tags { struct _primary_key_; struct _foreign_key_; struct _value_; } typedef fusion::map< fusion::vector< fields::_id_ , std::string, tags::_primary_key_ > , fusion::vector< fields::_name_ , std::string, tags::_value_ > , fusion::vector< fields::_birthday_ , std::string, tags::_value_ > , fusion::vector< fields::_address_id_, std::string, tags::_foreign_key_ > > table_type; int main() { /// how to get the primary key type vector } I was looking at fusion's filter_view but that is only available for non-metaprogramming. Any ideas? Christian

3 10

bind doesn't use tuple
by Gottlob Frege 22 Mar '07

22 Mar '07

why doesn't bind use tuple (internally)? I'm tempted to leave the email as just that. No more to say. OK, maybe one thing - I'm not asking because I necessarily think it should, I can imagine there are reasons, just wondering what they might be. Thanks Tony

4 5

Using Variant with StateChart events
by Federico J. Fernández 22 Mar '07

22 Mar '07

Hi Boosters! I have been trying to use those libraries together. To be concrete, I want to have a variant of boost::statechart::event derived classes. The problem is that when I try to compile it I receive a lot of error messages, but the important ones say that there is no match for "operator new".. I mean, variant is calling a prototype of operator new that event doesn't offer. This is the error that I get: /usr/include/boost/variant/variant.hpp: In constructor 'boost::variant<T0, T1, T2, T3, T4, T5, T6, T7, T8, T9, T10, T11, T12, T13, T14, T15, T16, T17, T18, T19>::variant() [with T0_ [...] /usr/include/boost/variant/variant.hpp:1174: error: no matching function for call to 'evt_reset::operator new(unsigned int, void*)' /usr/include/boost/statechart/event.hpp:38: note: candidates are: static void* boost::statechart::event<MostDerived, Allocator>::operator new(size_t) [with MostDerived = evt_reset, Allocator = std::allocator<void>] Any idea? I want to have a container of this "events" and don't want to have a base class and working with virtual functions. Thanks!

2 1

Statechart current status
by Philippe DAVID 22 Mar '07

22 Mar '07

Hi, I'm interested in using the statechart library and I would like to know what its current status is. It looks like it is inactive since dec 2005, am I right ? Has the project been freezed ? Btw, why is statechart not included with the other boost libs ? Thanks, Philippe

2 1

[Serialization, XML] - how to avoid another encoding level
by Frank Bergemann 22 Mar '07

22 Mar '07

Hi, i am working with MICO and use boost::serialization (XML) for import/export of CORBA functions IN/OUT parameters, because i have an XML formatted configuration for the data to use for CORBA service invocations. One of the problems was, that some of the MICO are template'd. E.g. a data type holding a number of digits (e.g. telephone #no) is encoded in MICO as: template <class T, int TID, int max> class BoundedSequenceTmpl { ... } ... using basic <class T> = CORBA::Short for digits (e.g.). 1st i used ToString(...)/fromString(...) functions to be able to save&load such to/from boost archives. And this worked - though it had the little drawback, that i needed to use additional variables in load() & save() to handle it this way. But then next it came up, that some other CORBA data types use the BoundedSequenceTempl<> as well, but for _more complex_ <class T> - e.g. BoundedSequenceTempl<> of BoundedSequenceTempl<>. For that i created a version of serialize(), save() & load() like this: template<class Archive, class T, int TID, int max> void serialize(Archive & ar, BoundedSequenceTmpl<T, TID, max> &p, const unsigned int version) { split_free(ar, p, version); } template<class Archive,class T, int TID, int max> void save(Archive & ar, const BoundedSequenceTmpl<T, TID, max> &p, const unsigned int version) { unsigned long length = p.length(); ar & make_nvp("length", length); for (unsigned long i = 0; i < length; ++i) { std::ostringstream name; name << "item_" << i; ar & make_nvp(name.str(), p[i]); } } template<class Archive, class T, int TID, int max> void load(Archive & ar, BoundedSequenceTmpl<T, TID, max> &p, const unsigned int version) { unsigned long length; ar & make_nvp("length", length); p.length(length); for (unsigned long i = 0; i < length; ++i) { std::ostringstream name; name << "item_" << i; ar & make_nvp(name.str(), p[i]); } } then i thought i can get rid of the formerly used ToString(...)/FromString(...) helpers and replace them by overloaded versions of the new (more generic) load() & save() variants dedicated for this BoundedSequenceTempl<...>. Like this: #if 0 // This was intended to make obsolete MicoVector_to_string(...) & // MicoVector_from_string(...) // kind of "template specialization" for the ones above // for BoundedSequenceTmpl<...> of types, which we don't want // to be spread out any further // // But i couldn't make this work. // Means: it compiles. But it implies another XML encoding level // - which is not what i want template<class Archive, int TID, int max> void serialize(Archive & ar, BoundedSequenceTmpl<CORBA::Short, TID, max> &p, const unsigned int version) { split_free(ar, p, version); } template<class Archive, int TID, int max> void save (Archive & ar, const BoundedSequenceTmpl<CORBA::Short, TID, max> &p, const unsigned int version) { std::ostringstream ostream; for (unsigned long i=0; i<p.length(); ++i) { ostream << p[i]; } // tried this, but it didn't work (compilation failed): ar & str; // so i tested this option std::string str(ostream.str()); ar & make_nvp("hallo", str); // CREATES ANOTHER ENCODING LEVEL :-( } template<class Archive, int TID, int max> void load (Archive & ar, BoundedSequenceTmpl<CORBA::Short, TID, max> &p, const unsigned int version) { ... } #endif But that created me another level of encoding - which i don't want. E.g. having a superior encoder: template<class Archive> void save(Archive & ar, const DeleteHlrSubscriber_s &pInstance, const unsigned int version) { std::string simsi; std::string smsisdn; MicoVector_to_string(pInstance.imsi, simsi); MicoVector_to_string(pInstance.msisdn, smsisdn); // ar & make_nvp("imsi", simsi); // replace by the next... // this worked, but see above (the by "#if 0" disabled stuff) ar & make_nvp("imsi", pInstance.imsi); ar & make_nvp("msisdn", smsisdn); ar & make_nvp( "attach_imsi", pInstance.attach_imsi); } using pInstance.imsi directly (which is a BoundedSequenceTmpl<CORBA::Short>()) resulted in: <imsi> // dropped boost::serializations internal XML attributes here <hallo>SomeString</hallo> </imsi> But I would like to have this one <imsi>SomeString</imsi> How can i do that? Thanks! Best regards! Frank _______________________________________________________________ SMS schreiben mit WEB.DE FreeMail - einfach, schnell und kostenguenstig. Jetzt gleich testen! http://f.web.de/?mc=021192

2 3