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

Some questions about using BOOST_CLASS_EXPORT_KEY macro of boost.serialization library
by ZouLei 26 Feb '10

26 Feb '10

Dear all , The library version I used was boost_1_42, and the IDE was VS2008 . When I attempt to use the BOOST_CLASS_EXPORT_KEY macro, I met some problems. In my word, I think it doesn't work as expected . My project files are listed following (just five tiny files) : // Base.h #ifndef _BASE_H #define _BASE_H #include <boost/serialization/access.hpp> #include <boost/serialization/export.hpp> #include <boost/serialization/type_info_implementation.hpp> #include <boost/serialization/extended_type_info_no_rtti.hpp> #include <boost/serialization/base_object.hpp> #include <boost/serialization/shared_ptr.hpp> class Base { public: Base() ; Base(int a) ; virtual ~Base(); private: friend class boost::serialization::access ; template<class Archive> void serialize(Archive& ar, const unsigned int version) { ar & a_ ; } private: int a_ ; }; BOOST_CLASS_EXPORT_KEY(Base) BOOST_SERIALIZATION_SHARED_PTR(Base) #endif //_BASE_H // Base.cpp #include "Base.h" Base::Base(): a_(0){ } Base::Base( int a ): a_(a){ } Base::~Base(){ } BOOST_CLASS_EXPORT_IMPLEMENT(Base) //Derived.h #ifndef _DERIVED_H #define _DERIVED_H #include "base.h" class Derived : public Base { public: Derived(); Derived(int a, int b); ~Derived(); private: friend class boost::serialization::access ; template<class Archive> void serialize(Archive& ar, const unsigned int version) { ar & boost::serialization::base_object<Base>(*this) ; ar & b_ ; } private: int b_ ; }; BOOST_CLASS_EXPORT_KEY(Derived) BOOST_SERIALIZATION_SHARED_PTR(Derived) #endif // _DERIVED_H // Drived.cpp #include "StdAfx.h" #include "Derived.h" Derived::Derived(): b_(0){ } Derived::Derived( int a, int b ): Base(a), b_(b){ } Derived::~Derived(){ } BOOST_CLASS_EXPORT_IMPLEMENT(Derived) // main.cpp #include <fstream> #include <iostream> #include <boost/archive/text_oarchive.hpp> #include <boost/serialization/shared_ptr.hpp> #include "Derived.h" int _tmain(int argc, _TCHAR* argv[]) { boost::shared_ptr<Base> base_ptr(new Derived(10,15)) ; std::ofstream ofs("test.txt") ; boost::archive::text_oarchive oa(ofs) ; oa << base_ptr ; return 0; } There are no errors in compile-time. But when I executed the program, a memory error caused ! I don't know what is wrong about my program. Whether my usage of BOOST_CLASS_EXPORT_KEY/BOOST_CLASS_EXPORT_IMPLEMENT was wrong ? And, how can I fix the problems ? Thanks, and with my best wishes. Lei

4 4

[Statechart] Question regarding state_machine event handling
by Arne Babnik 26 Feb '10

26 Feb '10

Hi, I'm currently working with the the statechart library in version 1.42.0 and have some difficulties with the way the simple state_machine processes its events. For my work I need several independent state machines, which communicate with each other using process_event/post_event. For various reasons I do not want to use the asynchroneous_state_machine - mostly because my application is built around a global event/message handler and I do not want to use multiple threads with their necessary synchronization. It basically works because in my hierarchy state_machine<>::process_event() is called only in one direction, while in the "opposite" direction (or if unsure) state_machine<>::post_event() is being used. E.g. A calls B->process_event(), B calls C->process_event(), but C calls B->post_event() and B calls A->post_event(). This works so far, and since the process_event() function also processes events posted from within its execution context I do not see any problems with this. However, what I could really make use of are two functions: a) a query function to check if there are any events in the state machine queue (if additional post_event() calls have been made outside the process_event() context); and b) a process_event() function that does not need an event argument, but only processes the queue. Currently I use a "dummy" event for this, but this is rather ugly. Does anybody know why such functions are not available? For b) there is a function named "process_queued_events()" that does exactly what I need, but unfortunately this function is declared private.For a), a simple "eventQueue_.empty()" wrapper would suffice. An even bigger problem for me is that process_event() executes the event argument _first_. I'm rather used to state machines that process their events in-order regardless of the way they got queued up, so I had expected that process_event() first executes the events already in the queue before processing the given event (in other words, the given event argument is only put to the end of the queue initially, and then the queue is processed). Any ideas why the state machine works this way? What is the rationale for this behaviour? Thanks, Arne

3 4

Re: [Boost-users] 1.39 serialisation: buggy warning message for shared_ptr
by Avi Bahra 26 Feb '10

26 Feb '10

Robert Ramey wrote >> The only reason to export a type is if one needs to >> serialize it through a base class pointer. So, to me >> the best way to suppress the warning would be >> to not export the type. The warning message appears even if you don't export the type: Here is a simple test case //================================== #include <boost/archive/text_iarchive.hpp> #include <boost/archive/text_oarchive.hpp> #include <boost/serialization/base_object.hpp> #include <boost/serialization/string.hpp> #include <boost/serialization/vector.hpp> #include <boost/serialization/shared_ptr.hpp> // no need to include shared_ptr #include <boost/serialization/export.hpp> class Concrete { public: Concrete(const std::string& name ) : name_(name) {} Concrete() {} private: std::string name_; friend class boost::serialization::access; template<class Archive> void serialize(Archive& ar,const unsigned int){ ar & name_; } }; class Base { public: Base(const std::string& name ) : name_(name) {} Base() {} virtual ~Base(){} private: std::string name_; *std::vector<boost::shared_ptr<Concrete> > vec_*; friend class boost::serialization::access; template<class Archive> void serialize(Archive & ar, const unsigned int) { ar & name_; ar & vec_; } }; class Derived : public Base { public: Derived(const std::string& name ) : Base(name) {} Derived() {} private: friend class boost::serialization::access; template<class Archive> void serialize(Archive & ar, const unsigned int) { ar & boost::serialization::base_object<Base>(*this); } }; BOOST_CLASS_EXPORT(Derived); //================================== Notice that I have not exported Concrete, but the serialisation will spit out a warning message, about class Concrete. ... boost_1_39_0/boost/serialization/extended_type_info_typeid.hpp", line 88: warning #2414-D: delete of pointer to incomplete class BOOST_STATIC_WARNING(boost::is_polymorphic<T>::value); ^detected during: instantiation of "const boost::serialization::extended_type_info *boost::serialization::extended_type_info_typeid<T>::get_derived_extended_type_info(const T &) const [with T=Concrete]" at line 92 of "/scratch/ma/emos/ma0/hpia64/boost/boost_1_39_0/boost/archive/shared_ptr_helper.hpp" It looks like if we are serialising via a base ptr, and the base class is serialising concrete data member via a shared ptr, it has assumed that the type must be polymorphic. This does not seem right to me ? Best regards, Ta, Avi

3 2

Multiple key types map
by Michele Caini 26 Feb '10

26 Feb '10

Hi all, I'm looking for a map that has no matter about key types, so I can use such map storing int indexed objects and also, i.e., char indexed objects, as all-in-one container. It can be useful to have such type of key-type unaware containers, as we can find with other languages. What about existing one?

5 5

[program_options] allow_guessing issue
by Filip Konvička 26 Feb '10

26 Feb '10

Hi, I've just upgraded to boost 1.42 from 1.40, and it seems that the new feature 'allow_guessing', a default setting, breaks my existing code. I have two options: "lang" and "lang-dir", the first obviously a prefix of the other, and when my program is run with "--lang=en_US --lang-dir=.", the first argument is interpreted as "lang-dir". Even though the documentation says somewhere that one argument should not be a prefix of another with 'allow_guessing' on, this is a breaking change and should be put in a big red box somewhere on the front page :-) If the change is intended to be kept for the future, it would be wise to detect the problem by examining the options, rather than wait until another problem occurs (in my case, the runtime exception is 'multiple_options' when the user uses both options, and - much worse! - the option gets misinterpreted without a warning if he only uses the shorter option). Cheers, Filip

2 2

Property tree example debug_settings.cpp
by Mauricio Gomes 26 Feb '10

26 Feb '10

Hi, In http://www.boost.org/doc/libs/1_42_0/libs/property_tree/examples/debug_sett… FOREACH lines are commented out. <debug> <filename>debug.log</filename> <modules> <module>Finance</module> <module>Admin</module> <module>HR</module> </modules> <level>2</level> </debug> Then I compiled and tested the program and noticed it wrote only one value for the <modules> tag above. Here it is the output: <?xml version="1.0" encoding="utf-8" ?> - <#> <debug> <filename>debug.log</filename> <level>2</level> - <#> <modules> <module>HR</module> </modules> </debug What is the status of this library ? Is this a known bug or it is just the example that is not uptodate ? If the latter, can anyone provide a functional version of debug_settings.cpp ? Thanks ! -- Mauricio

2 1

[fusion] Introspection of fused objects?
by Rutger ter Borg 26 Feb '10

26 Feb '10

Suppose I have a function object created on the basis of a sequence of fused function objects, where I would like to do shuffle the call order on the basis of some criterion. template< typename SeqOfFused > struct merge_fused { merge_fused( SeqOfFused f ): m_functions( f ) {} typedef void result_type; template< int N, typename Seq > void operator()( Seq& seq ) { // normal dispatch call would be fusion::at_c<N>(m_functions)( seq ); if ( some case ) { store seq for later use } else { call appropriate fused object with parts of stored stuff } } SeqofFused m_functions; // some state / memory specific to this group of fused objects some_state m_state; }; I'm doing a static cast to a superclass to get rid of the int N template parameter, to make the merge_fused look exactly like a fused function object. This works fine. Storing stuff is what my question is about: is there any way to know the type of Seq from its Fused counterpart? Looking at the code, the answer is no, but perhaps there exists introspection code in fusion which I am unaware of? One solution would be to simply use static casting based on Seq in operator(), but then my question would be: is that safe? Thanks, Rutger

2 3

valgrind and boost.datetime
by Matwey V. Kornilov 26 Feb '10

26 Feb '10

Hi, I initialize time input facet in such way: new boost::posix_time::time_input_facet( std::string("%Y-%m-%d %H:%M: %S") ) and valgrind --track-origins=yes tells me: ==30317== Conditional jump or move depends on uninitialised value(s) ... ==30317== Uninitialised value was created by a stack allocation ==30317== at 0x806C486: std::vector<std::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::allocator<std::basic_string<char, std::char_traits<char>, std::allocator<char> > > > boost::date_time::gather_weekday_strings<char>(std::locale const&, bool) (strings_from_facet.hpp:84) Is there way to fix the problem? The problem is located within boost according to valgrind, but I assume that it can be produced by incorrect usage. Thanks gcc 4.4.1 boost 1.39.0

2 3

Strange iterator_traits of iterator_facade
by Swerts, Benjamin 26 Feb '10

26 Feb '10

Hello all, I have a class that is templated on an iterator type. As it only works with bidirectional and random access iterators, I thought I would impose that as follows: template<typename T> class A { A() { checkIterator(std::iterator_traits<T>::iterator_category()); }; template<typename Tag> void checkIterator( Tag ); void checkIterator( std::bidirectional_iterator_tag ) {}; void checkIterator( std::random_access_iterator_tag ) {}; void checkIterator( boost::random_access_traversal_tag ) {}; }; This works when T is a standard iterator but not when it's one of my iterator_facade subclasses. For example: class const_iterator : public iterator_facade<const_iterator, const B, boost::random_access_traversal_tag, B> When using this class with A<const_iterator> I get a linker error because the checkIterator function is called with: boost::detail::iterator_category_with_traversal<struct std::input_iterator_tag,struct boost::random_access_traversal_tag> I certainly cannot add a specialization of checkIterator with this tag to class A. Could anyone shed some light on this problem? I'm working with Boost 1.39.0 on MSVC2008SP1. Thanks a lot in advance! Greets, Ben

2 3

[boost-users][filesystem] some missing function
by Sascha Ochsenknecht 26 Feb '10

26 Feb '10

Hi, in Boost.Filesystem I can find a function: template <class Path> std::time_t last_write_time(const Path& p); which is corresponding to st_mtime (last modification time) of the stat structure. Within the stat structure, their are also st_atime and st_ctime defined, but there are no corresponding functions. Is there a special reason for that? Maybe not portabel? Couldn't they be added? E.g. template <class Path> std::time_t last_access_time(const Path& p); template <class Path> std::time_t last_status_change_time(const Path& p); Cheers, Sascha

1 0

understanding Xcode and Boost
by David Rangel 26 Feb '10

26 Feb '10

Hello, Thanks to Igor R for helping me get started with the compilation problems of the regex test program at http://www.boost.org/doc/libs/1_35_0/more/getting_started/unix-variants.htm… Xcode. I renamed all id occurrences to idd in the regex headers, but still get errors. Also, I don't understand why this is necessary for Xcode but not using the command line compilation suggested in the getting started guide. Is Xcode generally more problematic when using libraries like Boost or GSL? Is there a better recommended development environment than Xcode, or is it simple once the error is tracked down? Thanks again for any input. Message: 7 Date: Wed, 24 Feb 2010 23:52:22 +0200 From: Igor R <boost.lists(a)gmail.com> To: boost-users(a)lists.boost.org Subject: Re: [Boost-users] Xcode and Boost Message-ID: <cfe0a3cf1002241352v15440e32wc44f9885861e2efc(a)mail.gmail.com > Content-Type: text/plain; charset=UTF-8 > I am interested in using Xcode on OSX to develop some code, and have been > having some troubles compiling the regex example given at: > http://www.boost.org/doc/libs/1_35_0/more/getting_started/unix-variants.html > I can compile using the command line, but not in Xcode. Rename all the occurrences of local variables named "id" in the relevant regex headers.

3 4

Re: [Boost-users] result_of protocol
by Hicham Mouline 26 Feb '10

26 Feb '10

>---------------------------------------------------------------------------------- >struct extract_rule { > > template <typename Sig> struct result; > > template <typename Holder> >struct result< extract_rule(const Holder) > { > typedef const typename Holder::type& type; > }; > template <typename Holder> > typename result< extract_rule(const Holder) >::type > operator()( const Holder& h ) const > { > return h.rule; > } >}; >---------------------------------------------------------------------------------- >1> with >1> [ >1> Sig=extract_rule (const onevariantrule<parameters_t,0,boost::spirit::istream_iterator> ) >1> ] >The template onevariantrule<> is defined _before_ extract_rule. >------------------------------------------------------------------------------------------- >template <typename parameters_t, unsigned n, typename Iterator> >struct onevariantrule { > typedef .... type; >}; >------------------------------------------------------------------------------------------- >Holder matches onevariantrule<parameters_t,0,boost::spirit::istream_iterator>. >That template instantiation is indeed a class. Both specializations template <typename Holder> struct result< extract_rule(const Holder) > { }; and template <typename Holder> struct result< extract_rule(const Holder& ) > { ///// the reference was missing here, doh! typedef const typename Holder::type& type; }; were needed. Sorry for the spam,

1 0

Boost.ASIO "deasynchronisation"
by Igore Dmit. 26 Feb '10

26 Feb '10

Hello, I'm developing networking library that provides RTMP-like networking using Boost.ASIO library. My problem is that any synchronous-access function breaks into many little functions that do some part of job and invoke async operation with "next" function as a handler, it makes the code really ugly and obfuscating. E.g. void hadshake that reads N bytes from peer, responds with those bytes + same count of other bytes, reads N bytes back and checks that they are the same. It breaks into: void start_handshake_wait(); void handle_handshake_wait(const boost::system::error_code&, std::size_t, buffer_ptr &); void handle_handshake_wait_write_dup(..same as above..); void handle_handshake_wait_write_mine(...); void handle_handshake_wait_read_mine_check(...); Every function uses boost::bind, so for some quite simple operation a lot of code is produced. Receiving a chunk of data breaks into: void read_header_head(); // header size depends on first byte void parse_header_head(); void parse_header_body(); void parse_chunk_body(); The only idea I came to is to use some tricks like coroutines. (Sorry for big listings) template<class T> class Context : public boost::enable_shared_from_this< Context<T> > { public: typedef boost::shared_ptr<Context<T> > ptr; private: typedef boost::shared_ptr<T> objptr_t; typedef void (T::*funcptr_t) (const BSYSECode&, std::size_t, Context<T>::ptr, buffer_ptr); objptr_t objptr_; funcptr_t funcptr_; public: std::map< std::string, buffer_ptr > buffers; int state; Context(objptr_t op, funcptr_t fp) : objptr_(op), funcptr_(fp), buffers(), state(0) { } template<typename AsyncReadStream> void async_read(AsyncReadStream &s, buffer_ptr buf) { boost::asio::async_read(s, *buf, boost::bind(funcptr_, objptr_, BASIOPErr, BASIOPBytes, this->shared_from_this(), buf)); } template<typename AsyncWriteStream> void async_write(AsyncWriteStream &s, buffer_ptr buf) { boost::asio::async_write(s, *buf, boost::bind(funcptr_, objptr_, BASIOPErr, BASIOPBytes, this->shared_from_this(), buf)); } }; #define CTX_PROLOG(c) \ if(!c) \ return; \ \ if(c->state == 0) { \ #define CTX_READ(c, s, b) \ c->state=__LINE__; \ c->async_read(s, b); \ return; \ } else if(c->state == __LINE__) { \ #define CTX_WRITE(c, s, b) \ c->state=__LINE__; \ c->async_write(s, b); \ return; \ } else if(c->state == __LINE__) { \ #define CTX_EPILOG \ } buffer_ptr is actually a shared_ptr to buffer, it's used to get buffer back to pool as it goes out of scope. With all this handshake becomes something more readable (to me, but it still cryptic; again, sorry for posting that much): class Handshaker : public boost::enable_shared_from_this<Handshaker> { private: Connection::ptr conn_; public: typedef boost::shared_ptr<Handshaker> ptr; typedef Context<Handshaker> context; Handshaker(Connection::ptr conn) : conn_(conn) { } void shake() { context::ptr ctx(new context(shared_from_this(), &Handshaker::perform)); perform(BSYSECode(), 0, ctx, buffer_ptr()); } private: void perform(const BSYSECode& err, std::size_t sz, context::ptr ctx, buffer_ptr buf) { if(err) return; CTX_PROLOG(ctx); CTX_READ(ctx, conn_->socket(), conn_->pool()->allocate(HandshakeSize)); CTX_WRITE(ctx, conn_->socket(), buf); ctx->buffers["mybuf"] = conn_->pool()->allocate(HandshakeSize); CTX_WRITE(ctx, conn_->socket(), ctx->buffers["mybuf"]); CTX_READ(ctx, conn_->socket(), conn_->pool()->allocate(HandshakeSize)); buffer_ptr mybuf = ctx->buffers["mybuf"]; buffer_ptr hisbuf = buf; if(boost::asio::buffer_size(*mybuf) != boost::asio::buffer_size(*hisbuf)) return; if(memcmp( boost::asio::buffer_cast<void*>(*mybuf), boost::asio::buffer_cast<void*>(*hisbuf), boost::asio::buffer_size(*mybuf) )) return; CTX_EPILOG; } }; Finally, the question: is there any library that provides functionality like that? I'd like to stick with it cause it looks like my implementation is really bad, HUGE bunch of limitations on code written that way, cryptic look, usage of macro, any local variable has to resort in ctx to survive async call.

4 3

Re: [Boost-users] result_of protocol
by Hicham Mouline 26 Feb '10

26 Feb '10

-----Original Message----- >From: "Hicham Mouline" [hicham(a)mouline.org] >Date: 26/02/2010 09:40 AM >To: boost-users(a)lists.boost.org >Subject: Re: result_of protocol. >---------------------------------------------------------------------------------- >struct extract_rule { > > template <typename Sig> struct result; > > template <typename Holder> >struct result< extract_rule(const Holder) > { > typedef const typename Holder::type& type; > }; > template <typename Holder> > typename result< extract_rule(const Holder) >::type > operator()( const Holder& h ) const > { > return h.rule; > } >}; >---------------------------------------------------------------------------------- >1> with >1> [ >1> Sig=extract_rule (const onevariantrule<parameters_t,0,boost::spirit::istream_iterator> ) >1> ] >The template onevariantrule<> is defined _before_ extract_rule. >------------------------------------------------------------------------------------------- >template <typename parameters_t, unsigned n, typename Iterator> >struct onevariantrule { > typedef .... type; >}; >------------------------------------------------------------------------------------------- >Holder matches onevariantrule<parameters_t,0,boost::spirit::istream_iterator>. >That template instantiation is indeed a class. In the snippet above, I replaced Holder by what the compiler tells me it is in the error log, as ------------------------------------------------------------ struct extract_rule { template <typename Sig> struct result; template <typename Holder> struct result< extract_rule(const Holder) > { typedef const onevariantrule<parameters_t,0,boost::spirit::istream_iterator>::type& type; }; ------------------------------------------------------------ And this works. If msvc is able to see that Holder is that particular template instantiation in the error log, isn't it able to know then that Holder is a class? regards,

1 0

[graph] parallel named graph and redistribute
by Andreas Zilian 26 Feb '10

26 Feb '10

Hello boost.graph users, I would like to use the PBGL with named vertices and later redistribute the graph. Unfortunately, I cannot compile the attached code, which - following the documentation - should work, since redistribute() just needs a property_map containing the new rank id's. A minimal example is shown below and a more verbose in the attachment. The example on redistribution in the tests of course works, but uses no named vertices. Are there any users of the parallel BGL who managed to solve this issue? And how? Moreover, could the new (serial) labeled_graph be an option? Thank you very much, Andreas #include <boost/graph/use_mpi.hpp> #include <boost/graph/distributed/adjacency_list.hpp> #include <boost/graph/distributed/mpi_process_group.hpp> using namespace boost; using boost::graph::distributed::mpi_process_group; //----------------------------------------------------------------------------- struct Topo { friend class boost::serialization::access; template<class Archive> void serialize(Archive & ar, const unsigned int version) { ar & user_id & rank_id ; } Topo( int uid=0, int rid=-1 ) : user_id(uid), rank_id(rid) {} int user_id; int rank_id; friend std::ostream& operator<< (std::ostream& o, const Topo& t) { o << "uid = " << t.user_id << ", rid = " << t.rank_id; return o; } }; //----------------------------------------------------------------------------- // named vertices in a graph namespace boost { namespace graph { template<> struct internal_vertex_name< Topo > { typedef multi_index::member< Topo, int, &Topo::user_id > type; }; template<> struct internal_vertex_constructor< Topo > { typedef vertex_from_name< Topo > type; }; }} typedef adjacency_list< vecS, distributedS< mpi_process_group, vecS >, directedS, Topo > Graph; int main(int argc, char* argv[]) { boost::mpi::environment env(argc,argv); Graph g; property_map< Graph, int Topo::*>::type rid_map = get(&Topo::rank_id, g); g.redistribute( rid_map ); return 0; }

1 0

result_of protocol
by Hicham Mouline 26 Feb '10

26 Feb '10

Hello, I get the following msvc2008 error: error C2825: 'Holder': must be a class or namespace when followed by '::' in the following code ---------------------------------------------------------------------------------- struct extract_rule { template <typename Sig> struct result; template <typename Holder> struct result< extract_rule(const Holder) > { typedef const typename Holder::type& type; }; template <typename Holder> typename result< extract_rule(const Holder) >::type operator()( const Holder& h ) const { return h.rule; } }; ---------------------------------------------------------------------------------- 1> with 1> [ 1> Sig=extract_rule (const onevariantrule<parameters_t,0,boost::spirit::istream_iterator> ) 1> ] The template onevariantrule<> is defined _before_ extract_rule. ------------------------------------------------------------------------------------------- template <typename parameters_t, unsigned n, typename Iterator> struct onevariantrule { typedef .... type; }; ------------------------------------------------------------------------------------------- Holder matches onevariantrule<parameters_t,0,boost::spirit::istream_iterator>. That template instantiation is indeed a class. It has the inner field called "type". Any ideas would be appreciated,

1 0

for help, some error of boost regx
by pirate-king 26 Feb '10

26 Feb '10

Hello everyone, When I use boost_1_41_0 of regx in vs2005, I met this runtime error: boost::exception_detail::clone_impl<boost::exception_detail::error_info_injector<boost::regex_error> > I search a lot, but there is not valid method to solve it. Can you do me a favor to give me the way to solve it. Thanks a lot my code is #include <boost/regex.hpp> #include <iostream> boost::regex e("(\\d{3, 4})[- ]?(\\d{4)})[- ]?(\\d{4})[- ]?(\\d{4})"); void print_captures(const std::string& regx, const std::string& text) { boost::regex e(regx); boost::smatch what; std::cout << "Expression: \"" << regx << "\"\n"; std::cout << "Text: \"" << text << "\"\n"; if (boost::regex_match(text, what, e, boost::match_extra)) { unsigned i, j; std::cout << "** Match found **\n Sub-Expressions: \n"; for (i = 0; i < what.size(); ++i) std::cout << " $" << i << " = \"" << what[i] << "\"\n"; std::cout << " Captures:\n"; for (i = 0; i < what.size(); ++i) { std::cout << " $" << i << " = {"; for (j = 0; j < what.captures(i).size(); ++j) { if (j) std::cout << ", "; else std::cout << " "; std::cout << "\"" << what.captures(i)[j] << "\""; } std::cout << "}\n"; } } else { std::cout << "** No match found **\n"; } } int main(int , char* []) { print_captures("(([[:lower:]]+)|([[:upper:]]+))+", "aBBcccDDDDeeeeeee"); return 0; }

3 2

io_service run thread exits on network disconnect (chat server example)
by A.Agrawal＠patrick.com.au 26 Feb '10

26 Feb '10

Hello, My code is similar to the one in chat_server example in boost documentation. In my program I use boost asio to communicate with server. Client has following code to connect to the server: Class Members: class DJClient { tcp::socket *_socket; boost::asio::io_service io_service; boost::thread *t; ............................MORE........................................... } void DJClient::Connect() //To Connect to the server { tcp::resolver resolver(io_service); tcp::resolver::query query("10.21.7.5", "50000"); tcp::resolver::iterator endpoint_iterator = resolver.resolve(query); _socket = new tcp::socket(io_service); tcp::endpoint endpoint = *endpoint_iterator; _socket->async_connect(endpoint, boost::bind(&DJClient::handle_connect, this, boost::asio::placeholders::error, ++endpoint_iterator)); t = new boost::thread(boost::bind(&boost::asio::io_service::run, &io_service)); return; } //Callback function after connection has been attempted, also reads data asynchronously from the server void DJClient::handle_connect(const boost::system::error_code& error, tcp::resolver::iterator endpoint_iterator) { if (!error) { boost::asio::async_read(*_socket, boost::asio::buffer(_inbound_data_length), boost::bind(&DJClient::handle_read_data_length, this, boost::asio::placeholders::error)); } else if (endpoint_iterator != tcp::resolver::iterator()) { _socket->close(); tcp::endpoint endpoint = *endpoint_iterator; _socket->async_connect(endpoint, boost::bind(&DJClient::handle_connect, this, boost::asio::placeholders::error, ++endpoint_iterator)); } } Client writes to the server using the following code, this happens in the main thread (the one that called Connect), asynchronously void DJClient::GetAllWIs(const char* oow_id) { io_service.post(boost::bind(&DJClient::write_wis_for_oow_req, this, oow_id)); } void DJClient::write_wis_for_oow_req(const char* oow_id) { boost::asio::async_write(*_socket, oow_id, boost::bind(&DJClient::handle_write, this, boost::asio::placeholders::error)); } Everything works fine without any problem until network cable is unplugged. After network cable is unplugged, a call to GetAllWI does not call write_wis_for_oow_req. I can see an exit thread message in Visual Studio output window. I suspect that io_service::run exits and io_service is no longer valid and that is why it does not call the method. My program needs to handle this scenario and Client should be able to detect network problem and reconnect. 1. It would be great help if someone can suggest a way to handle this scenario. Is there a way to know this io_service::run has exited. 2. I can see in watch window that io_service._impl.stopped changes its value from 0 to 1. But I do not know how to access it programmatically. Any response will be greatly appreciated as I am not very experience in using boost. Thanks in advance Anil Agrawal Patrick Technology & Systems

2 1

[boost-program-options] Minor documentation bug
by mmocny 26 Feb '10

26 Feb '10

To whom it may concern, The last code example in the section "Option Details" on page http://www.boost.org/doc/libs/1_42_0/doc/html/program_options/tutorial.html, the line: $bin/gcc/debug/options_description ...should be bold, to match the rest of the syntax. I found the example confusing until I properly parsed what was meant to be written. Might I also suggest a space after the $ or even* "shell> command" * -Michal Mocny

3 2

[boost-users][array] Passing by reference
by Ryan McConnehey 26 Feb '10

26 Feb '10

I'm wanting to pass a boost::array by reference to a function. Normally, this isn't a problem. My first approach yielded a compiler error stating boost::array needs template parameters. When adding the template parameters (approach 2) the compiler then states boost::array needs T to be a compile time constant. I want the function to be able to take different boost::array sizes. Can boost::array be passed by reference without knowing the size of the array at compile time? Any help would be great. Ryan Approach 1 void extract (boost::array & var) {} Approach 2 template<typename T, typename S> void extract (boost::array<T, S> & var) {}

2 2

Boost.Serialization warning C4540 on MSVC
by Antti Alho 26 Feb '10

26 Feb '10

Hi, I'm using Boost.Serialization for my project and have run into a cryptic warning message. Although it doesn't seem to actually do any harm on run-time I would like to know what it means, and possibly how I can fix it without #pragma disabling it. The relevant code and the entire warning can be found here: Codepad paste <http://codepad.org/MVx2O6dC> The error message was produced by Visual Studio 2008. Since I'm not perfectly sure what the correct syntax is supposed to be regarding lines 74 and 75 when dealing with classes using multiple inheritance and since I couldn't find documentation regarding the subject those could very well be the source of my problems (although it seems to do what it is expected to do). All help greatly appreciated, thanks in advance.

3 2

[program_options][1.35] validating string value
by Soumen 26 Feb '10

26 Feb '10

Hi, My option (say, "opt1") can take any of only few possible string type values (say, "value1", "value2", "value3" and "value4"). Is there anyway to specify this such that if any other value, say - "valueXX", is passed to "opt1", it throws an error? If yes, can someone point me to documentation or sample code snippet to show how to do this? Or, explicitly, checking against these possible values is my only option? Regards, ~ Soumen -- View this message in context: http://old.nabble.com/-program_options--1.35--validating-string-value-tp276… Sent from the Boost - Users mailing list archive at Nabble.com.

2 1

Combing 2 memory pools
by Michael Kennard 26 Feb '10

26 Feb '10

Short version: I'm using the boost pool library for allocation of a linked list of small objects. Now I want to append one linked list onto the end of another. Thus I also want to combine the memory pools. I don't see this functionality in the interface documentation. Is there a way to easily combine two memory pools? Longer version: The list is a member of a class which manages it and has methods for creating and deleting list elements. The memory pool is also a member of this class. Before using the memory pool, each list element was allocated individually and deleted individually. With the pool, deletion is easier because you just destroy the pool instead of deleting each list element individually. Merging two lists together before I used the pool was just a matter of appending one linked list onto the end of the other (easy, and code not shown). But now that I'm using the pool I've got to figure out how to merge the two pools into one. Here's a simplified version of the code (which I haven't tried to compile): struct element { int m_data; element* m_next; }; class MyList { public: MyList(); // constructor void NewElement (int data) { element* new_element = static_cast<element*>(m_pool.malloc()); new_element->m_data = data; // add to beginning of list new_element->next = m_list; m_list = new_element; } element* m_list; boost::pool m_pool; }; -- Michael Kennard Aquaveo (www.aquaveo.com)

2 1

fusion Transformation function passed to transform
by Hicham Mouline 26 Feb '10

26 Feb '10

Hello, The transformation function to pass to fusion::transform() function and metafunction are hard to get right, in particular the embedded inner result struct because of the fact that the template argument of result is a "function signature" There is only the 1 simple example in libs/fusion/doc/html/fusion/algorithm/transformation/metafunctions/transform.html I for one would love to see another example where the fusion vector has different types, and where the result of transformation function depends on each type of the sequence transform is applied on. In the meantime, I am writing something like this: struct extract_rule { template <typename Signature> struct result; //// primary template template <typename Holder> struct result< extract_rule(const Holder) > { //// is this always a specialization of the primary? typedef .... type; }; template <typename Holder> typename result<extract_rule(const Holder)>::type operator()( const Holder& h ) const { .... } }; Does this look right? regards,

2 1

function pointers and anys
by Alan Tennant 25 Feb '10

25 Feb '10

I created a pointer to a function, I am able to call the function though the pointer as well as move the pointer, as I would expect. I put it into a boost::any. How do I get the pointer out of the any so that I can call it? I'm having difficulty using the any_cast to do it.

2 2